prompt stringlengths 5.5k 11.7k | env_class stringclasses 1
value | reward_spec stringlengths 131 9k | extra_info stringlengths 2 48 |
|---|---|---|---|
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Support for this work came from multiple sources. The HKU provided grants GJTD-2020-09, GJTD-2019-11, and GJTD-2018-01 to Petersen and Ivanov, who express their sincere gratitude; the ORAU also contributed funding. Petersen acknowledges the Agencia Nacional de Investigación e Innovación for support received during the 2005 Fall semester. Partial funding for Kim was secured through NSF CMMI academic research grant CMMI 1760916. Hospitality provided by the Ministero dell'Istruzione, dell'Università e della Ricerca in 2010, the HUS in Munich in 2012, and the United States Department of Energy Office of Science in 2015 during final project stages is deeply appreciated. We are thankful for financial backing from the HKU Strategic Research Grant (HKU grant GJTD-2020-01). The United States Department of Energy Office of Science deserves special recognition from Petersen and Kim as the location where most final work was completed during Spring 2015."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "HKU", "awards": [{"funding_scheme": ["Strategic Research Grant"], "award_ids": ["GJTD-2020-09", "GJTD-2019-11", "GJTD-2018-01", "GJTD-2020-01"], "award_title": []}]}, {"funder_name": "ORAU", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Agencia Nacional de Investigación e Innovación", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "NSF CMMI", "awards": [{"funding_scheme": [], "award_ids": ["CMMI 1760916"], "award_title": []}]}, {"funder_name": "Ministero dell'Istruzione, dell'Università e della Ricerca", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "HUS", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "United States Department of Energy Office of Science", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "The study was supported by the Universiti Sains Malaysia (USM) Research University Grant (1001/CIPPT/8012203)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Universiti Sains Malaysia (USM)", "awards": [{"funding_scheme": [], "award_ids": ["1001/CIPPT/8012203"], "award_title": []}]}]} | {"doi": "10.21203/rs.3.rs-49230/v3"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Support came from MINECO and federal sources via the Alzheimer's Association (Agreement 235789). Publication content does not represent the Alzheimer's Association policy positions. Commercial entity references carry no endorsement. | E.K. has funding from the Rockefeller Foundation (2017 YWS 307 GA-S), the Science and Technology Department of Hubei Province, and UW (NFRFT-2022-00197). V.M. and B.G. are funded by the Ministry of Health and Welfare. MINECO supported development and upkeep of the online portal. T.R. receives salary through MINECO."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "MINECO", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Alzheimer's Association", "awards": [{"funding_scheme": [], "award_ids": ["235789"], "award_title": []}]}, {"funder_name": "Rockefeller Foundation", "awards": [{"funding_scheme": [], "award_ids": ["2017 YWS 307 GA-S"], "award_title": []}]}, {"funder_name": "Science and Technology Department of Hubei Province", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "UW", "awards": [{"funding_scheme": [], "award_ids": ["NFRFT-2022-00197"], "award_title": []}]}, {"funder_name": "Ministry of Health and Welfare", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This research was funded by the National Natural Science Foundation of China General Program. The authors are grateful to N. Okonkwo and F. Schmidt and J. Petrov for valuable input."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "National Natural Science Foundation of China", "awards": [{"funding_scheme": ["General Program"], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Funding: The costs associated with open-access article-processing charges (APCs) have been underwritten by UMP, a program available to authors at non-commercial entities throughout New Zealand. | Acknowledgment: We are grateful to UMP, whose financial backing enabled the publication of this manuscript in an open-access format."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "UMP", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Portions of the experimental work were conducted at Universität Hannover under the auspices of NWO Open Competition for Domain Science grant 438.19.902. Financial support for author one included: NWO (Grant No. 057-14-002, awarded to Patricia O'Connor) during the CMU period; The Polish Academy of Sciences awards 2015/19/B/ST4/02721 together with 2015/19/D/HS4/02574 for the \"Designer Microstructure via Optimal Transport Theory\" initiative (awarded to Sebastian Park) during the Polish Academy of Sciences tenure. Author two's research received partial funding via NWO (Grant Nos. 731.015.410, ALWOP.395) at CMU, via the National Institute for Health Research support programs (RP-PG-0612-20004 as well as NIHR202024), and via the National Alliance for Research on Schizophrenia and Depression (award NARSAD Independent Award 2015). The authors wish to thank Henrik Larsson for constructive feedback."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Universität Hannover", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "NWO", "awards": [{"funding_scheme": ["Open Competition for Domain Science"], "award_ids": ["438.19.902", "057-14-002", "731.015.410", "ALWOP.395"], "award_title": []}]}, {"funder_name": "Polish Academy of Sciences", "awards": [{"funding_scheme": [], "award_ids": ["2015/19/B/ST4/02721", "2015/19/D/HS4/02574"], "award_title": []}]}, {"funder_name": "National Institute for Health Research", "awards": [{"funding_scheme": [], "award_ids": ["RP-PG-0612-20004", "NIHR202024"], "award_title": []}]}, {"funder_name": "National Alliance for Research on Schizophrenia and Depression", "awards": [{"funding_scheme": [], "award_ids": ["NARSAD Independent Award 2015"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Financial support came from SANCST under grant 14-BIO1978-03. Principal investigators: M. Chen and L. Zhang."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "SANCST", "awards": [{"funding_scheme": [], "award_ids": ["14-BIO1978-03"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Funding: This work was supported by the priority program SPP2225 (FU 1027/4-1 to M.F.), the DFG (SU 239/18-1 to R.S.) and the Collaborative Research Centers CRC1449, Project ID 431232613; project B5 (to M.F. and F.G.). J.K. is financed by a scholarship from the H. Wilhelm Schaumann foundation. M.F. received support by the Freie Universität Berlin within the Excellence Initiative of the German Research Foundation. JGH is funded by a National Science Centre (NCN, Poland) grant no. 2020/39/B/NZI/02898 (OPUS 20)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "DFG", "awards": [{"funding_scheme": [], "award_ids": ["SU 239/18-1"], "award_title": []}]}, {"funder_name": "Collaborative Research Centers CRC1449", "awards": [{"funding_scheme": [], "award_ids": ["431232613"], "award_title": []}]}, {"funder_name": "H. Wilhelm Schaumann foundation", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Freie Universität Berlin", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "German Research Foundation", "awards": [{"funding_scheme": ["Excellence Initiative"], "award_ids": [], "award_title": []}]}, {"funder_name": "National Science Centre (NCN, Poland)", "awards": [{"funding_scheme": ["OPUS 20"], "award_ids": ["2020/39/B/NZI/02898 (OPUS 20)"], "award_title": []}]}, {"funder_name": null, "awards": [{"funding_scheme": ["Priority program SPP2225"], "award_ids": ["FU 1027/4-1"], "award_title": []}]}]} | {"doi": "10.1101/2022.09.15.507240"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Financial support was provided by multiple agencies: the Collaborative Research Excellence Grant grant TED2021-129770B-C21, Petróleo Brasileiro S.A. awards 5900.0118828.21.2 and 465389/2014-7, the Frontier Research Excellence Grant under 306631, and the Chancellor's Fellowship - University of Edinburgh grant 833925 from University of Edinburgh."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": null, "awards": [{"funding_scheme": ["Collaborative Research Excellence Grant"], "award_ids": ["TED2021-129770B-C21"], "award_title": []}]}, {"funder_name": "Petróleo Brasileiro S.A.", "awards": [{"funding_scheme": [], "award_ids": ["5900.0118828.21.2", "465389/2014-7"], "award_title": []}]}, {"funder_name": null, "awards": [{"funding_scheme": ["Frontier Research Excellence Grant"], "award_ids": ["306631"], "award_title": []}]}, {"funder_name": "University of Edinburgh", "awards": [{"funding_scheme": ["Chancellor's Fellowship"], "award_ids": ["833925"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "The authors acknowledge Junta de Castilla y León, the Department of Atomic Energy (Centre Digitisation.Bavaria, ZD.B), SHMEC, and Queensland University of Technology for partial support. Special thanks to Lars Nielsen for manuscript review and constructive feedback."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Junta de Castilla y León", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Department of Atomic Energy", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "SHMEC", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Queensland University of Technology", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Acknowledgements. This research was conducted during the periods the second author was a graduate student at Princeton University and was employed by the Clay Mathematics Institute as a Liftoff Fellow. The second author was partially supported by a Graduate School Centennial Fellowship at Princeton University. The third author is partially supported by grant No.10631060 of the National Natural Science Foundation of China."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Clay Mathematics Institute", "awards": [{"funding_scheme": ["Liftoff Fellow"], "award_ids": [], "award_title": []}]}, {"funder_name": "Princeton University", "awards": [{"funding_scheme": ["Graduate School Centennial Fellowship"], "award_ids": [], "award_title": []}]}, {"funder_name": "National Natural Science Foundation of China", "awards": [{"funding_scheme": [], "award_ids": ["10631060"], "award_title": []}]}]} | {"doi": "10.48550/arxiv.0801.1943"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Funding for this research was provided by the ESRC Secondary Data Analysis Initative, Grant Number ES/K004433/1."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "ESRC Secondary Data Analysis Initative", "awards": [{"funding_scheme": [], "award_ids": ["ES/K004433/1"], "award_title": []}]}]} | {"doi": "10.48550/arxiv.1411.2514"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "S. Williams (rkonda@ucsb.edu), T. Anderson, B. Zhang and J. P. Hassan are with the Department of Electrical and M. University of Cambridge, G. Silva, CA. The present study is in part supported by Varian Medical Systems (United States) Grant #R01EB032680 and Region of Sardinia Grant #LR7/2007."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Varian Medical Systems (United States)", "awards": [{"funding_scheme": [], "award_ids": ["R01EB032680"], "award_title": []}]}, {"funder_name": "Region of Sardinia", "awards": [{"funding_scheme": [], "award_ids": ["LR7/2007"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "JME - supported by the National Institutes of Health (NIH) through Grant Numbers R01HD079076 and R03HD094577: Eunice Kennedy Shriver National Institute of Child Health & Human Development; National Center for Medical Rehabilitation Research. DMW – data for the SYMBIOME project were supported by a Canadian Pain Society Early Career Investigator Award, an Ontario Ministry of Research and Innovation Early Researcher Award, and funding from the Strategy for Patient-Oriented Research (SPOR) program for chronic pain jointly funded by the Canadian Institutes of Health Research and matching industry partners. All funders were arm’s length and had no influence on the interpretation or reporting of results. | Joy MacDermid was supported by a Canadian Institutes of Health Research Chair in Gender, Work and Health and the Dr. James Roth Chair in Musculoskeletal Measurement and Knowledge Translation."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "National Institutes of Health (NIH)", "awards": [{"funding_scheme": [], "award_ids": ["R01HD079076", "R03HD094577"], "award_title": []}]}, {"funder_name": "Canadian Pain Society", "awards": [{"funding_scheme": [], "award_ids": ["Early Career Investigator Award"], "award_title": []}]}, {"funder_name": "Ontario Ministry of Research and Innovation", "awards": [{"funding_scheme": [], "award_ids": ["Early Researcher Award"], "award_title": []}]}, {"funder_name": "Canadian Institutes of Health Research", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": ["Chair in Gender, Work and Health"]}, {"funding_scheme": [], "award_ids": [], "award_title": ["Dr. James Roth Chair in Musculoskeletal Measurement and Knowledge Translation"]}]}, {"funder_name": null, "awards": [{"funding_scheme": ["Strategy for Patient-Oriented Research (SPOR) program"], "award_ids": [], "award_title": []}]}]} | {"doi": "10.21203/rs.3.rs-16949/v1"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Funding for this study came from NSF DBI (project 1457541) with additional backing from McGill University."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "NSF DBI", "awards": [{"funding_scheme": [], "award_ids": ["1457541"], "award_title": []}]}, {"funder_name": "McGill University", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Funding was provided by the Commission européenne under the Seventh Framework Programme award 608322. | Cytometric work occurred at the Single Cell Analysis Facility of the Research Sciences Complex with the Ninth Institute of Second Ministry of Mechanical Industry funding. | L.W. receives a Shanghai Municipal Education Commission fellowship (20142BAB217008). | N.S. was supported by the Scottish Government Strategic Research Excellence Grant program."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Commission européenne", "awards": [{"funding_scheme": ["Seventh Framework Programme"], "award_ids": ["608322"], "award_title": []}]}, {"funder_name": "Ninth Institute of Second Ministry of Mechanical Industry", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Shanghai Municipal Education Commission", "awards": [{"funding_scheme": [], "award_ids": ["20142BAB217008"], "award_title": []}]}, {"funder_name": "Scottish Government", "awards": [{"funding_scheme": ["Strategic Research Excellence Grant"], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Universidad de Misuri provided financial support for this research. M.T. was supported through salary funding from Jiāngsū Dàxué."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Universidad de Misuri", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Jiāngsū Dàxué", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Financial backing was obtained from Xīnán Kējìdàxué grant 61379005 and from PSU award COVID19-CE-2020-53 through NREL subcontract 934521. We appreciate valuable exchanges with Priya Patel, Manuel Silva, Anastasia Kozlov, and Wei Zhang."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Xīnán Kējìdàxué", "awards": [{"funding_scheme": [], "award_ids": ["61379005"], "award_title": []}]}, {"funder_name": "PSU", "awards": [{"funding_scheme": [], "award_ids": ["COVID19-CE-2020-53"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This research is sponsored by grants from the Laboratory Directed Research and Development Program of Oak Ridge National Laboratory (ORNL), managed by UT-Battelle, LLC for the U. S. Department of Energy under Contract No. DE-AC05-00OR22725 and the Natural Sciences and Engineering Research Council of Canada and Canadian Foundation for Innovation."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Oak Ridge National Laboratory (ORNL)", "awards": [{"funding_scheme": ["Laboratory Directed Research and Development Program"], "award_ids": ["DE-AC05-00OR22725"], "award_title": []}]}, {"funder_name": "UT-Battelle, LLC", "awards": [{"funding_scheme": [], "award_ids": ["DE-AC05-00OR22725"], "award_title": []}]}, {"funder_name": "U. S. Department of Energy", "awards": [{"funding_scheme": [], "award_ids": ["DE-AC05-00OR22725"], "award_title": []}]}, {"funder_name": "Natural Sciences and Engineering Research Council of Canada", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Canadian Foundation for Innovation", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {"doi": "10.48550/arxiv.0706.3546"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "*We acknowledge partial funding from the Science and Technology Department of Qinghai Province (2021-ZJ-906) and the Beijing Municipal Education Commission (KJZD20191002302) for this research."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Science and Technology Department of Qinghai Province", "awards": [{"funding_scheme": [], "award_ids": ["2021-ZJ-906"], "award_title": []}]}, {"funder_name": "Beijing Municipal Education Commission", "awards": [{"funding_scheme": [], "award_ids": ["KJZD20191002302"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "The authors acknowledge Mr. Alan Hoofring (NIH’s Medical Arts Design Section) for help with Illustrations (Figures 1 and 6). In addition, the authors thank Ms. Sherly Michel (NIDCD) and Mr. Pat Diers (NIDCD) for help with animal care and genotyping and Dr. Kuni Iwasa (NIDCD) and Dr. Benjamin Perrin (Indiana University - Purdue University Indianapolis) for carefully reading the manuscript and providing critical inputs. A.X.C.R. and this work was supported by the National Institutes of Health (NIH) Distinguished Scholars Program and the NIH Intramural Research Program of the National Institute of Biomedical Imaging and Bioengineering (grant # ZIA EB000094). This research was supported (in part) by the Intramural Research Program of the NIH, NIDCD DC000039 to T.B.F. S.K. was supported by JSPS KAKENHI grant 20K09687 and G.F. by NIDCD/NIH (R01DC014658 and S10OD025130). The electron microscopy was performed at the University of Kentucky Electron Microscopy Center, which belongs to the National Science Foundation NNCI Kentucky Multiscale Manufacturing and Nano Integration Node, supported by ECCS-1542174."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "National Institutes of Health (NIH)", "awards": [{"funding_scheme": ["Distinguished Scholars Program"], "award_ids": [], "award_title": []}, {"funding_scheme": ["Intramural Research Program"], "award_ids": [], "award_title": []}]}, {"funder_name": "National Institute of Biomedical Imaging and Bioengineering", "awards": [{"funding_scheme": ["NIH Intramural Research Program"], "award_ids": ["ZIA EB000094"], "award_title": []}]}, {"funder_name": "NIDCD", "awards": [{"funding_scheme": [], "award_ids": ["DC000039", "R01DC014658", "S10OD025130"], "award_title": []}]}, {"funder_name": "JSPS", "awards": [{"funding_scheme": ["KAKENHI"], "award_ids": ["20K09687"], "award_title": []}]}, {"funder_name": "National Science Foundation", "awards": [{"funding_scheme": ["NNCI"], "award_ids": ["ECCS-1542174"], "award_title": ["NNCI Kentucky Multiscale Manufacturing and Nano Integration Node"]}]}]} | {"doi": "10.1101/2021.06.28.450090"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This work receives financial backing from the David and Lucile Packard Foundation under the Fellowships in Science and Engineering program."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "David and Lucile Packard Foundation", "awards": [{"funding_scheme": ["Fellowships in Science and Engineering"], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Participants and peer reviewers are thanked for enabling this research. Financial resources came from China University of Geosciences (Beijing) Fundamental Research Funds for the Central Universities (2022170, CUG110813), the United States Department of Energy Office of Nuclear Energy Nuclear Fuel Cycle and Supply Chain (UIDB/04516/2020,PTDC/CCI-INF/32081/2017, SFRH/BPD/100377/2014), the National Natural Science Foundation of China General Program (TH-201918), along with Clínica Mayo Translational Neuroscience Research Grant (CA77118). Data were partially obtained from the Cognitive Mapping Project (CMP), which receives Franklin College support under UG1025GR182025 and Victoria University of Manchester funding."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "China University of Geosciences (Beijing)", "awards": [{"funding_scheme": ["Fundamental Research Funds for the Central Universities"], "award_ids": ["2022170", "CUG110813"], "award_title": []}]}, {"funder_name": "United States Department of Energy Office of Nuclear Energy Nuclear Fuel Cycle and Supply Chain", "awards": [{"funding_scheme": [], "award_ids": ["UIDB/04516/2020,PTDC/CCI-INF/32081/2017", "SFRH/BPD/100377/2014"], "award_title": []}]}, {"funder_name": "National Natural Science Foundation of China", "awards": [{"funding_scheme": ["General Program"], "award_ids": ["TH-201918"], "award_title": []}]}, {"funder_name": "Clínica Mayo", "awards": [{"funding_scheme": ["Translational Neuroscience Research Grant"], "award_ids": ["CA77118"], "award_title": []}]}, {"funder_name": "Franklin College", "awards": [{"funding_scheme": [], "award_ids": ["UG1025GR182025"], "award_title": []}]}, {"funder_name": "Victoria University of Manchester", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Essential resources for LifeCourse are provided by UNISA, MCI via grant 300391FRB17PAGAN/99/2021, and the Danish Ministry of Higher Education and Science. A detailed grant list can be found on the LifeCourse portal. NSF IOS provides support to JMA. A postdoctoral award from the Chongqing Science and Technology Commission grant 2021XM3016 supports REW. CMAH's work is funded by a Advanced Research Fellowship Program. JMA serves as guarantor for this publication authored by the research team."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "UNISA", "awards": [{"funding_scheme": [], "award_ids": ["300391FRB17PAGAN"], "award_title": []}]}, {"funder_name": "MCI", "awards": [{"funding_scheme": [], "award_ids": ["99/2021"], "award_title": []}]}, {"funder_name": "Danish Ministry of Higher Education and Science", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "NSF IOS", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Chongqing Science and Technology Commission", "awards": [{"funding_scheme": [], "award_ids": ["2021XM3016"], "award_title": []}]}, {"funder_name": null, "awards": [{"funding_scheme": ["Advanced Research Fellowship Program"], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Funding for this research came from UCLM and General Electric (United Kingdom) (U01DA051039)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "UCLM", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "General Electric (United Kingdom)", "awards": [{"funding_scheme": [], "award_ids": ["U01DA051039"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "*Kyoto University, Japan. The Jiangsu Provincial Department of Education Award KYCX22_0901, Ministerstvo Zemědělství České Republiky grant #MZE-RO 1523, and a Division of Atmospheric and Geospace Sciences grant provided partial funding. Email: h.tanaka@kyoto-u.ac.jp | †University of Cambridge, UK. The Tianjin Municipal Science and Technology Commission grants 18YFZCGX00360, 24ZYCGSY00570 and a Simons Investigator Award supported this work. Email: as@cam.ac.uk"}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Jiangsu Provincial Department of Education", "awards": [{"funding_scheme": [], "award_ids": ["KYCX22_0901"], "award_title": []}]}, {"funder_name": "Ministerstvo Zemědělství České Republiky", "awards": [{"funding_scheme": [], "award_ids": ["MZE-RO 1523"], "award_title": []}]}, {"funder_name": "Division of Atmospheric and Geospace Sciences", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Tianjin Municipal Science and Technology Commission", "awards": [{"funding_scheme": [], "award_ids": ["18YFZCGX00360", "24ZYCGSY00570"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This project was supported by AB INNOV under Grant No. CHRPJ 462510. Additional funding was provided by the Collaborative Health Research Partnership Grant under Grant No. AB INNOV."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "AB INNOV", "awards": [{"funding_scheme": ["Collaborative Health Research Partnership Grant"], "award_ids": ["CHRPJ 462510"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Acknowledgments. This research was supported in part by the National Natural Science Foundation of China (62172246), the Open Project Program of State Key Laboratory of Virtual Reality Technology and Systems (VRLAB2021A05), and Youth Innovation and Technology Support Plan of Colleges and Universities in Shandong Province (2021KJ062)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "National Natural Science Foundation of China", "awards": [{"funding_scheme": [], "award_ids": ["62172246"], "award_title": []}]}, {"funder_name": "State Key Laboratory of Virtual Reality Technology and Systems", "awards": [{"funding_scheme": ["Open Project Program"], "award_ids": ["VRLAB2021A05"], "award_title": []}]}, {"funder_name": "Colleges and Universities in Shandong Province", "awards": [{"funding_scheme": ["Youth Innovation and Technology Support Plan"], "award_ids": ["2021KJ062"], "award_title": []}]}]} | {"doi": "10.48550/arxiv.2206.09564"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This work received support from Leona M. and Harry B. Helmsley Charitable Trust grant P30DK043351 and the Estonian Research Council the Personal the Research Project Grant for G.H.N. Grant DH/DSH2008 from the Department of Health and Social Care and the National Institute on Alcohol Abuse and Alcoholism Research Project Grant R01 DK113196 funded R.W., I.B., and O.Z. Leona M. and Harry B. Helmsley Charitable Trust grant AWD-007061 provided support to K.F. and Y.S."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Leona M. and Harry B. Helmsley Charitable Trust", "awards": [{"funding_scheme": [], "award_ids": ["P30DK043351", "AWD-007061"], "award_title": []}]}, {"funder_name": "Estonian Research Council", "awards": [{"funding_scheme": ["Personal Research Project Grant"], "award_ids": [], "award_title": []}]}, {"funder_name": "Department of Health and Social Care", "awards": [{"funding_scheme": [], "award_ids": ["DH/DSH2008"], "award_title": []}]}, {"funder_name": "National Institute on Alcohol Abuse and Alcoholism", "awards": [{"funding_scheme": ["Research Project Grant"], "award_ids": ["R01 DK113196"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "The National Natural Science Foundation of China provided multiple awards to P.S.: the General Program 2016A610097 and 2010A610015, General Program 202003N4025, plus the General Program 2019A60100,2019A610101. K.S. received the Division of Cancer Epidemiology and Genetics at the National Cancer Institute Research Project Grant Ro1CA 214913 and Julius-Maximilians-Universität Würzburg Bätz Prize Bätz Preis 2024."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "National Natural Science Foundation of China", "awards": [{"funding_scheme": ["General Program", "General Program", "General Program"], "award_ids": ["2016A610097", "2010A610015", "202003N4025", "2019A60100,2019A610101"], "award_title": []}]}, {"funder_name": "Division of Cancer Epidemiology and Genetics at the National Cancer Institute", "awards": [{"funding_scheme": ["Research Project Grant"], "award_ids": ["Ro1CA 214913"], "award_title": []}]}, {"funder_name": "Julius-Maximilians-Universität Würzburg", "awards": [{"funding_scheme": ["Bätz Prize"], "award_ids": ["Bätz Preis 2024"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "The research contributions of F. Yamamoto and C. L. Kim were funded by the Strategic Research Advancement Grant under award 2024C018-4. This manuscript presents work undertaken jointly with the National Office of Philosophy and Social Sciences Major Project of Social Science Research."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": null, "awards": [{"funding_scheme": ["Strategic Research Advancement Grant"], "award_ids": ["2024C018-4"], "award_title": []}]}, {"funder_name": "National Office of Philosophy and Social Sciences", "awards": [{"funding_scheme": ["Major Project of Social Science Research"], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "We acknowledge Martin Fischer for helpful guidance regarding computational methods and GAUSSIAN software utilization. Financial backing came from Nanyang Technological University through the \"Advanced materials — Laboratory to application\" funding initiative via the collaborative project PHOTONIX (RG24/05)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Nanyang Technological University", "awards": [{"funding_scheme": [], "award_ids": ["RG24/05"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "The researcher is grateful to KU Strategic Vision Research Grant, Research to Prevent Blindness Career Development Award, the Ministry of Health Labour and Welfare, the Ministry of Health Labour and Welfare, the National Natural Science Foundation of China, Majmaah University, Research to Prevent Blindness, and Nihon Gakujutsu Shinkō Kai."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "KU", "awards": [{"funding_scheme": ["Strategic Vision Research Grant"], "award_ids": [], "award_title": []}]}, {"funder_name": "Research to Prevent Blindness", "awards": [{"funding_scheme": ["Career Development Award"], "award_ids": [], "award_title": []}]}, {"funder_name": "Ministry of Health Labour and Welfare", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "National Natural Science Foundation of China", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Majmaah University", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Nihon Gakujutsu Shinkō Kai", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "The authors acknowledge the Beijing Municipal Administration of Hospitals for support via the Clinical Medicine Development Project."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Beijing Municipal Administration of Hospitals", "awards": [{"funding_scheme": ["Clinical Medicine Development Project"], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This project received funding under Grant No. 20-3212-1 from KIGAM."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "KIGAM", "awards": [{"funding_scheme": [], "award_ids": ["20-3212-1"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "We gratefully acknowledge Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro Apoio ao Desenvolvimento Científico e Tecnológico em Projetos de Pesquisa for funding."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro", "awards": [{"funding_scheme": ["Apoio ao Desenvolvimento Científico e Tecnológico em Projetos de Pesquisa"], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Funding was provided through awards: 962-805808, RGPIN-2023-04100."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": null, "awards": [{"funding_scheme": [], "award_ids": ["962-805808", "RGPIN-2023-04100"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This research received funding from USYD under award DP110101740. The funding organization had no involvement in research design, data gathering and evaluation, or manuscript writing."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "USYD", "awards": [{"funding_scheme": [], "award_ids": ["DP110101740"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This work was funded by the National Cancer Research Institute, grant identifier LCSuPaC 30. Staff at the CRCTU who contributed to the trial were supported by a core funding grant (C22436/A25354)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "National Cancer Research Institute", "awards": [{"funding_scheme": [], "award_ids": ["C22436/A25354", "LCSuPaC 30"], "award_title": []}]}]} | {"doi": "10.21203/rs.3.rs-753663/v1"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Banque Mondiale provided the first author with partial support via the Global Partnership for Social Accountability Grant grant TF014205-CN. Grant R03-AA014598 from the National Institute on Alcohol Abuse and Alcoholism partially supported the second author. The third author benefited from combined financing by the Centros para el Control y Prevención de Enfermedades Public Health Preparedness and Response Initiative and Italian national funds administered through the University of Arizona Strategic Research Seed Grant Research Funding Program the Faculty Seed Grant for Cross-Institutional Research advancing cross-institutional research efforts."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Banque Mondiale", "awards": [{"funding_scheme": ["Global Partnership for Social Accountability Grant"], "award_ids": ["TF014205-CN"], "award_title": []}]}, {"funder_name": "National Institute on Alcohol Abuse and Alcoholism", "awards": [{"funding_scheme": [], "award_ids": ["R03-AA014598"], "award_title": []}]}, {"funder_name": "Centros para el Control y Prevención de Enfermedades", "awards": [{"funding_scheme": ["Public Health Preparedness and Response Initiative"], "award_ids": [], "award_title": []}]}, {"funder_name": "University of Arizona", "awards": [{"funding_scheme": ["Strategic Research Seed Grant", "Faculty Seed Grant for Cross-Institutional Research"], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This study was supported by the National Natural Science Foundation of China (82060551; 82060666)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "National Natural Science Foundation of China", "awards": [{"funding_scheme": [], "award_ids": ["82060551", "82060666"], "award_title": []}]}]} | {"doi": "10.21203/rs.3.rs-4099955/v1"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Acknowledgments This work was started at Dubna in August of 2009, when both of us were participating at Karolinska Institute on University of Cambridge and École Polytechnique are grateful to Université de Montréal for discussions during this Heidelberg University also wish to thank B. SilvaBraun and University of Copenhagen for useful communications prior and after this event. One of us (ALK) is grateful to his colleagues for creating rather stimulating atmosphere during his stay at Tsinghua University of CERN up to 24 October, 2009 and to the Members of University of São Paulo of RADCOR-2009 for the invitation, hospitality and financial support during his stay at this important Peking University work of both of us was supported in part by the Stellenbosch University grant No. OCE 0622825, while the work of ALK in part by the grant of Univerzita Karlova v Praze UNCE/SCI/13 as well."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Stellenbosch University", "awards": [{"funding_scheme": [], "award_ids": ["OCE 0622825"], "award_title": []}]}, {"funder_name": "Univerzita Karlova v Praze", "awards": [{"funding_scheme": [], "award_ids": ["UNCE/SCI/13"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Funding came from multiple programs: Ministerstvo zdravotnictví České republiky Internal Grant Agency of the Ministry of Health IP6003/0064203, Ministerstvo zdravotnictví České republiky Agency for Healthcare Research Development Project AZV 15-28998A, Chinese Academy of Fishery Sciences Central Public-interest Scientific Institution Basal Research Fund 2023TD56, and the Chinese Academy of Fishery Sciences Central Public-interest Scientific Institution Basal Research Fund 2021SD03."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Ministerstvo zdravotnictví České republiky", "awards": [{"funding_scheme": ["Internal Grant Agency of the Ministry of Health", "Agency for Healthcare Research Development Project"], "award_ids": ["IP6003/0064203", "AZV 15-28998A"], "award_title": []}]}, {"funder_name": "Chinese Academy of Fishery Sciences", "awards": [{"funding_scheme": ["Central Public-interest Scientific Institution Basal Research Fund", "Central Public-interest Scientific Institution Basal Research Fund"], "award_ids": ["2023TD56", "2021SD03"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "The authors gratefully acknowledge the support provided by National Natural Science Foundation of China (CN)(Grant nos. 21007060) and the Ji Yang college of Zhejiang Agriculture and Forestry University Science developmental foundation (Grant nos.4251700015) and Natural Science Foundation of Zhejiang Province (CN)(Grant nos. LY14B070014)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "National Natural Science Foundation of China (CN)", "awards": [{"funding_scheme": [], "award_ids": ["21007060"], "award_title": []}]}, {"funder_name": "Ji Yang college of Zhejiang Agriculture and Forestry University", "awards": [{"funding_scheme": ["Science developmental foundation"], "award_ids": ["4251700015"], "award_title": []}]}, {"funder_name": "Natural Science Foundation of Zhejiang Province (CN)", "awards": [{"funding_scheme": [], "award_ids": ["LY14B070014"], "award_title": []}]}]} | {"doi": "10.21203/rs.3.rs-1094738/v1"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Financial resources were obtained from the Jiangsu Provincial Department of Education in addition to several ASU grants including FA9550-15-1-0196, TG-DMR170070 (T.E.H.), SFI-04/BR/B0577 (P.H.T.), XH51027 (T.E.H., P.H.T.), the NIH Training Grant 12/RC/2289 (M.C.W.H.), and AZ51289 (M.W.)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Jiangsu Provincial Department of Education", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "ASU", "awards": [{"funding_scheme": [], "award_ids": ["FA9550-15-1-0196", "TG-DMR170070", "SFI-04/BR/B0577", "XH51027", "12/RC/2289", "AZ51289"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Funding for this research came from Comisión Sectorial de Investigación Científica (Proyectos de I+D), UNSW Sydney and the Division of Molecular & Cellular Biosciences."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Comisión Sectorial de Investigación Científica", "awards": [{"funding_scheme": ["Proyectos de I+D"], "award_ids": [], "award_title": []}]}, {"funder_name": "UNSW Sydney", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Division of Molecular & Cellular Biosciences", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Directorate for Biological Sciences provided support to J.M.B. The Ministry of Trade, Industry and Energy funded both A.K.S. and D.F.L. for their work on MS-related investigations."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Directorate for Biological Sciences", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Ministry of Trade, Industry and Energy", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "1.3. Acknowledgements. We are grateful to Uri Erez, Swastik Kopparty, and Alex Samorodnitsky for useful discussions. The authors gratefully acknowledge the support of grants ISF 2919/19, ISF 1791/17, BSF 2016256, the Simons Collaboration on Algorithms and Geometry, a Simons Investigator Award, and by the National Science Foundation (NSF) under Grant No. CCF-1814524."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "ISF", "awards": [{"funding_scheme": [], "award_ids": ["2919/19", "1791/17"], "award_title": []}]}, {"funder_name": "BSF", "awards": [{"funding_scheme": [], "award_ids": ["2016256"], "award_title": []}]}, {"funder_name": "National Science Foundation (NSF)", "awards": [{"funding_scheme": [], "award_ids": ["CCF-1814524"], "award_title": []}]}, {"funder_name": "Simons", "awards": [{"funding_scheme": ["Simons Collaboration on Algorithms and Geometry"], "award_ids": [], "award_title": []}, {"funding_scheme": ["Simons Investigator Award"], "award_ids": [], "award_title": []}]}]} | {"doi": "10.48550/arxiv.2006.00340"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "NMP was funded by the William Harvey Research Foundation, HPR and FRMBO were funded by National Council for Scientific and Technological Development (CNPq) fellowship. This study was supported by the German Research Foundation to LM (DFG, MA 7082/3-1), to CS (DFG, STO 1099/8-1) and by an intramural grant to EZ (START 131/19), National Council for Scientific and Technological Development to RS (CNPq, Brazil, Grant 409018/2018-0)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "William Harvey Research Foundation", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "National Council for Scientific and Technological Development (CNPq)", "awards": [{"funding_scheme": [], "award_ids": ["409018/2018-0"], "award_title": []}]}, {"funder_name": "German Research Foundation", "awards": [{"funding_scheme": [], "award_ids": ["MA 7082/3-1", "STO 1099/8-1", "START 131/19"], "award_title": []}]}]} | {"doi": "10.1101/2021.04.28.21255719"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "We acknowledge funds awarded to E. Mueller from the Department of Health (HI17C0104) and the Institute of Information & Communications Technology Planning & Evaluation. R. Martinez was the recipient of the Internal Postdoctoral Fellowship awarded by KU Leuven. E. Mueller holds the Education Department of Hunan Province Provincial University Key Research Project."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Department of Health", "awards": [{"funding_scheme": [], "award_ids": ["HI17C0104"], "award_title": []}]}, {"funder_name": "Institute of Information & Communications Technology Planning & Evaluation", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "KU Leuven", "awards": [{"funding_scheme": ["Internal Postdoctoral Fellowship"], "award_ids": [], "award_title": []}]}, {"funder_name": "Education Department of Hunan Province", "awards": [{"funding_scheme": ["Provincial University Key Research Project"], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "We gratefully acknowledge Universidad de California en Davis for project 71971123 that supported this research."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Universidad de California en Davis", "awards": [{"funding_scheme": [], "award_ids": ["71971123"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Financial backing came from Semiconductor Research Corporation via award number 2021-AH-3022. This investigation took place while A. Martinez was a visitor at TU Delft."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Semiconductor Research Corporation", "awards": [{"funding_scheme": [], "award_ids": ["2021-AH-3022"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Brown University provided funding for this research through award UM1 CA176726."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Brown University", "awards": [{"funding_scheme": [], "award_ids": ["UM1 CA176726"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "The author would like to thank V. Nguyen for pointing out the question considered here and for the many discussions, I. Rossi for valuable comments that helped improving the presentation of the article, as well as R. Patel for useful conversations. The author would also like to thank the anonymous referees for pointing our an error in the previous draft of the article and their valuable comments that greatly helped to improve the presentation. This work received support from the March of Dimes Birth Defects Foundation [5-FY13-208]. The Centers for Disease Control and Prevention."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "March of Dimes Birth Defects Foundation", "awards": [{"funding_scheme": [], "award_ids": ["5-FY13-208"], "award_title": []}]}, {"funder_name": "Centers for Disease Control and Prevention", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Funding was provided through awards: IT15715. We acknowledge B. Santos, R. Patel and S. Singh for their contributions."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": null, "awards": [{"funding_scheme": [], "award_ids": ["IT15715"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "The authors would like to thank Pratik Chaudari for valuable conversations as well as funding by NSF Grant No. DGE-1321851 and the Intel Science and Technology Center for Wireless Autonomous Systems. Any opinions, findings, and conclusions do not necessarily reflect the views of the NSF."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "NSF", "awards": [{"funding_scheme": [], "award_ids": ["DGE-1321851"], "award_title": []}]}, {"funder_name": "Intel Science and Technology Center for Wireless Autonomous Systems", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {"doi": "10.48550/arxiv.1902.06862"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This work is supported by the National Science and Engineering Research Council of Canada (NSERC), discovery grant program, and by the ETS Research Chair on Artificial Intelligence in Medical Imaging."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "National Science and Engineering Research Council of Canada (NSERC)", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "ETS", "awards": [{"funding_scheme": ["ETS Research Chair on Artificial Intelligence in Medical Imaging"], "award_ids": [], "award_title": []}]}]} | {"doi": "10.48550/arxiv.1811.08305"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Funding Information: This research was supported by a NARSAD Young Investigator Grant and a National Institute of Mental Health R21 grant (R21-MH112925) to Dr. Strauss."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "NARSAD", "awards": [{"funding_scheme": ["Young Investigator Grant"], "award_ids": [], "award_title": []}]}, {"funder_name": "National Institute of Mental Health", "awards": [{"funding_scheme": ["R21"], "award_ids": ["R21-MH112925"], "award_title": []}]}]} | {"doi": "10.31234/osf.io/c9sw4"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "SW was supported by the Royal Society of Edinburgh (RSE) (grant number 69938). BRH and JW acknowledge the receipt of studentship awards from the Health Data Research UK-The Alan Turing Institute Wellcome PhD Programme in Health Data Science (Grant Ref: 218529/Z/19/Z)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Royal Society of Edinburgh (RSE)", "awards": [{"funding_scheme": [], "award_ids": ["69938"], "award_title": []}]}, {"funder_name": "Health Data Research UK", "awards": [{"funding_scheme": ["Health Data Research UK-The Alan Turing Institute Wellcome PhD Programme in Health Data Science"], "award_ids": ["218529/Z/19/Z"], "award_title": []}]}]} | {"doi": "10.48550/arxiv.2209.15321"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Funding was obtained from the Département de l'Agriculture des États-Unis Agricultural Research Service National Program Grant through agreement USDA ARS Project 6040-32000-084-000D for HORIZON (horizon-systems.eu)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Département de l'Agriculture des États-Unis", "awards": [{"funding_scheme": ["Agricultural Research Service National Program Grant"], "award_ids": ["USDA ARS Project 6040-32000-084-000D"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This work was done while AM was at Microsoft Research.\nAM acknowledges the concurrent support in part by a\ngrant from the Open Philanthropy Project to the Center\nfor Human-Compatible AI, and in part by NSF grant\nCAREER IIS-1452099."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Open Philanthropy Project", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "NSF", "awards": [{"funding_scheme": ["CAREER"], "award_ids": ["IIS-1452099"], "award_title": []}]}]} | {"doi": "10.48550/arxiv.1905.05179"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Funding was obtained from the Shanghai Education Development Foundation via the Education Innovation and Talent Development Grant, as well as the National Institute of Justice, as well as NSF CHE."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Shanghai Education Development Foundation", "awards": [{"funding_scheme": ["Education Innovation and Talent Development Grant"], "award_ids": [], "award_title": []}]}, {"funder_name": "National Institute of Justice", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "NSF CHE", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This material is based upon work supported by the Ministry of Natural Resources under Ludwig Maximilian University GCMAC1907. | This report was prepared as an account of work sponsored by an agency of the Tsinghua University the University of Melbourne nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the University of Melbourne or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the University of Melbourne or any agency thereof."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Ministry of Natural Resources", "awards": [{"funding_scheme": [], "award_ids": ["GCMAC1907"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Ch.F. was partly supported by the US-Israel Binational Science Foundation, grant number 2014055, AFOSR, grant DMS-1265524, and NSF, grant FA9550- 12-1-0425. S.I. was partly supported RFBR, grant 20-01-00070, M.L. was supported by Academy of Finland, grants 273979 and 284715, and H.N. was partly supported by NSF grant DMS-1620102 and a Ramanujan Fellowship and a Swarna Jayanti fellowship."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "US-Israel Binational Science Foundation", "awards": [{"funding_scheme": [], "award_ids": ["2014055"], "award_title": []}]}, {"funder_name": "AFOSR", "awards": [{"funding_scheme": [], "award_ids": ["DMS-1265524"], "award_title": []}]}, {"funder_name": "NSF", "awards": [{"funding_scheme": [], "award_ids": ["FA9550- 12-1-0425", "DMS-1620102"], "award_title": []}]}, {"funder_name": "RFBR", "awards": [{"funding_scheme": [], "award_ids": ["20-01-00070"], "award_title": []}]}, {"funder_name": "Academy of Finland", "awards": [{"funding_scheme": [], "award_ids": ["273979", "284715"], "award_title": []}]}, {"funder_name": null, "awards": [{"funding_scheme": ["Ramanujan Fellowship"], "award_ids": [], "award_title": []}]}, {"funder_name": null, "awards": [{"funding_scheme": ["Swarna Jayanti fellowship"], "award_ids": [], "award_title": []}]}]} | {"doi": "10.48550/arxiv.1910.05084"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "We would like to thank Prof. Tom Clemente and the University of Nebraska Plant Transformation Core Research Facility for transformations, Erik Nielsen for providing the Golgi localization reporter, Gregg Sobocinski for imaging advice and Zach Lippman for providing the fin4 seeds. This work was supported by the National Science Foundation under grant no. IOS-1755482."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "National Science Foundation", "awards": [{"funding_scheme": [], "award_ids": ["IOS-1755482"], "award_title": []}]}]} | {"doi": "10.1101/2022.08.16.504199"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This work was partially supported by the H2020 project PRECRIME, funded under the ERC Advanced Grant 2017 Program (ERC Grant Agreement n. 787703). The driving simulator has been provided by BeamNG GmbH."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "ERC", "awards": [{"funding_scheme": ["H2020"], "award_ids": ["787703"], "award_title": ["PRECRIME"]}]}, {"funder_name": "BeamNG GmbH", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {"doi": "10.48550/arxiv.2107.06997"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This study was supported by funding from the Key Project of the National Natural Science Foundation of China (U21A20390), Jilin Scientific and Technological Development Program (20230204075YY, 20240305034YY), and the Graduate Innovation Program of Jilin University (2024CX280)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "National Natural Science Foundation of China", "awards": [{"funding_scheme": ["Key Project of the National Natural Science Foundation of China"], "award_ids": ["U21A20390"], "award_title": []}]}, {"funder_name": null, "awards": [{"funding_scheme": ["Jilin Scientific and Technological Development Program"], "award_ids": ["20230204075YY", "20240305034YY"], "award_title": []}]}, {"funder_name": "Jilin University", "awards": [{"funding_scheme": ["Graduate Innovation Program"], "award_ids": ["2024CX280"], "award_title": []}]}]} | {"doi": "10.21203/rs.3.rs-4513090/v1"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "*Corresponding author. † The first two authors contributed equally to this work. This work was partially supported by NSFC (62301432,62306240), NSBRPS (2023-JC-QN-0685, QCYRCXM-2023-057)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "NSFC", "awards": [{"funding_scheme": [], "award_ids": ["62301432", "62306240"], "award_title": []}]}, {"funder_name": "NSBRPS", "awards": [{"funding_scheme": [], "award_ids": ["2023-JC-QN-0685", "QCYRCXM-2023-057"], "award_title": []}]}]} | {"doi": "10.48550/arxiv.2404.14132"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "We acknowledge funding for the primary researcher from NSF DMR award 2128632 and Compagnia di San Paolo award Progetti di Ricerca di Ateneo 2012. Prifysgol Leeds grant 201506120055 supported the second researcher. The third researcher was funded in part by the Rockefeller Foundation grant 101046041 and Prifysgol Leeds grant 110 University Scholarship."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "NSF DMR", "awards": [{"funding_scheme": [], "award_ids": ["2128632"], "award_title": []}]}, {"funder_name": "Compagnia di San Paolo", "awards": [{"funding_scheme": [], "award_ids": ["Progetti di Ricerca di Ateneo 2012"], "award_title": []}]}, {"funder_name": "Prifysgol Leeds", "awards": [{"funding_scheme": [], "award_ids": ["201506120055", "110 University Scholarship"], "award_title": []}]}, {"funder_name": "Rockefeller Foundation", "awards": [{"funding_scheme": [], "award_ids": ["101046041"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This work was supported by the NSF and AFOSR."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "NSF", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "AFOSR", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {"doi": "10.48550/arxiv.1205.0767"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "The present study acknowledges support from AICTE (8-T4/FDC/RPS (Policy-l) /2019-2020), (8-196/RIFD/RPS(POLICY-1)/2018-19), (2-28/IKS Center-2/2022-23/54), the Advanced Research Development Grant (ZYGX2016J127), (25GZ0103212), Université Concordia (V01189); finally, King Mongkut's University of Technology North Bangkok (KMUTNB-FF-66-46)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "AICTE", "awards": [{"funding_scheme": [], "award_ids": ["8-T4/FDC/RPS (Policy-l) /2019-2020", "8-196/RIFD/RPS(POLICY-1)/2018-19", "2-28/IKS Center-2/2022-23/54"], "award_title": []}]}, {"funder_name": null, "awards": [{"funding_scheme": ["Advanced Research Development Grant"], "award_ids": ["ZYGX2016J127", "25GZ0103212"], "award_title": []}]}, {"funder_name": "Université Concordia", "awards": [{"funding_scheme": [], "award_ids": ["V01189"], "award_title": []}]}, {"funder_name": "King Mongkut's University of Technology North Bangkok", "awards": [{"funding_scheme": [], "award_ids": ["KMUTNB-FF-66-46"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "The authors thank University of Palermo ; additionally the Division of Ocean Sciences (OCE-1559178271) ; additionally King Mongkut's University of Technology Thonburi for funding."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "University of Palermo", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Division of Ocean Sciences", "awards": [{"funding_scheme": [], "award_ids": ["OCE-1559178271"], "award_title": []}]}, {"funder_name": "King Mongkut's University of Technology Thonburi", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This project received funding from ANU via the Competitive Research Grant, 973302, along with support from WHO (IVD/17/08)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "ANU", "awards": [{"funding_scheme": ["Competitive Research Grant"], "award_ids": ["973302"], "award_title": []}]}, {"funder_name": "WHO", "awards": [{"funding_scheme": [], "award_ids": ["IVD/17/08"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This work was funded by the German Research Foundation (DFG) via SFB 1333 project number 358283783 and the cluster of excellence EXC 2075 (SimTech) project number 390740016. Furthermore the authors acknowledge support by the state of Baden-Württemberg through bwHPC."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "German Research Foundation (DFG)", "awards": [{"funding_scheme": [], "award_ids": ["SFB 1333 project number 358283783", "EXC 2075 (SimTech) project number 390740016"], "award_title": []}]}, {"funder_name": "state of Baden-Württemberg", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "bwHPC", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {"doi": "10.26434/chemrxiv-2021-39nhv-v3"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "We thank the reviewer for constructive feedback. Partial funding for D.W.K. came from A*STAR together with a Severo Ochoa Postdoctoral Fellowship at Guelph. M.N. received partial support via a Gobierno del Principado de Asturias Severo Ochoa Postdoctoral Fellowship, the National Natural Science Foundation of China, and the National Agency for Research and Development. A*STAR partially funded R.P. M.R holds a postdoctoral position in Pure Mathematics at Nanjing University of Information Science and Technology."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "A*STAR", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Gobierno del Principado de Asturias", "awards": [{"funding_scheme": ["Severo Ochoa Postdoctoral Fellowship", "Severo Ochoa Postdoctoral Fellowship"], "award_ids": [], "award_title": []}]}, {"funder_name": "National Natural Science Foundation of China", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "National Agency for Research and Development", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Nanjing University of Information Science and Technology", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Suranaree University of Technology supported this work | Funding was provided by Intel Grant No. MOST 104-2221-E-011-165 | Support came from the National Institute of Allergy and Infectious Diseases Grant No. R01AI137418 | Funding was provided by Grant No. AI50698 | Funding was provided by Grant No. R01 AI174526 | Funding was provided by the Research Project Grant."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Suranaree University of Technology", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Intel", "awards": [{"funding_scheme": [], "award_ids": ["MOST 104-2221-E-011-165"], "award_title": []}]}, {"funder_name": "National Institute of Allergy and Infectious Diseases", "awards": [{"funding_scheme": ["Research Project Grant"], "award_ids": ["R01AI137418", "AI50698", "R01 AI174526"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "We acknowledge the Sigrid Jusélius Foundation Senior Investigator Grant (003701165704 to H. Lee), the Sigrid Jusélius Foundation (220196 to W. Johnson), and the Hebei Science and Technology Department project ID 246Z7738G, project E2 (to H. Lee and I. Kowalski). G. Alvarez is supported by a CDST scholarship. H. Lee received A*STAR support through the Sigrid Jusélius Foundation Senior Medical Investigator Grant. Y. Ahmed is funded by the British Council Newton International Fellowship grant 127407379."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Sigrid Jusélius Foundation", "awards": [{"funding_scheme": ["Senior Investigator Grant", "Senior Medical Investigator Grant"], "award_ids": ["003701165704", "220196"], "award_title": []}]}, {"funder_name": "Hebei Science and Technology Department", "awards": [{"funding_scheme": [], "award_ids": ["246Z7738G"], "award_title": []}]}, {"funder_name": "CDST", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "A*STAR", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "British Council", "awards": [{"funding_scheme": ["Newton International Fellowship"], "award_ids": ["127407379"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This work was supported by National Natural Science Foundation of China (No. 82072076, 82072165 and 81871556), and Young Top Talents in Xing Liao Talents Program of Liaoning Province (XL YC1807183)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "National Natural Science Foundation of China", "awards": [{"funding_scheme": [], "award_ids": ["82072076", "82072165", "81871556"], "award_title": []}]}, {"funder_name": "Liaoning Province", "awards": [{"funding_scheme": ["Young Top Talents in Xing Liao Talents Program of Liaoning Province"], "award_ids": ["XL YC1807183"], "award_title": []}]}]} | {"doi": "10.21203/rs.3.rs-871716/v1"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "The authors gratefully acknowledge funding from the Gordon and Betty Moore Foundation through the Science and Innovation Award, grant 9075."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Gordon and Betty Moore Foundation", "awards": [{"funding_scheme": [], "award_ids": ["9075"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This work is supported by CME Group. We thank them for providing us with valuable data and we are grateful for their donation of the Maxeler FPGA hardware. We especially acknowledge Ryan Eavy, Executive Director, Architectures at CME Group for his support and introduction to Maxeler."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "CME Group", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {"doi": "10.48550/arxiv.1706.01833"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "The Celgene Innovative Research Grant Program provided support under project 133826 (award Celgene Research Award 2017). The Institute of Applied Biology and Molecular Systems managed Celgene Investigator Initiated Research Award grant 27297. The State Science and Technology Commission award 2023ZD0506205 supported this study. Strains were maintained through Materials Research Science and Engineering Centers funding (DMR-2104841)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Celgene", "awards": [{"funding_scheme": ["Innovative Research Grant Program", "Investigator Initiated Research Award"], "award_ids": ["133826", "Celgene Research Award 2017", "27297"], "award_title": []}]}, {"funder_name": "State Science and Technology Commission", "awards": [{"funding_scheme": [], "award_ids": ["2023ZD0506205"], "award_title": []}]}, {"funder_name": "Materials Research Science and Engineering Centers", "awards": [{"funding_scheme": [], "award_ids": ["DMR-2104841"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "The UGC DAE Consortium for Scientific Research awarded grant CSR-IC-ISUM-26/CRS-309 to M.J.C., who also received support from the State Education Commission Scientific Research Project Fund. T.X., Q.V., and U.P. acknowledge funding from Xunta de Galicia grant EQC2019-005892-P and the Education Department of Hunan Province Key Research Project of Higher Education Institutions 202006. UGC DAE Consortium for Scientific Research award CSR-IC/CRS-89 supported H.N. and W.R."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "UGC DAE Consortium for Scientific Research", "awards": [{"funding_scheme": [], "award_ids": ["CSR-IC-ISUM-26/CRS-309", "CSR-IC/CRS-89"], "award_title": []}]}, {"funder_name": "State Education Commission", "awards": [{"funding_scheme": ["Scientific Research Project Fund"], "award_ids": [], "award_title": []}]}, {"funder_name": "Xunta de Galicia", "awards": [{"funding_scheme": [], "award_ids": ["EQC2019-005892-P"], "award_title": []}]}, {"funder_name": "Education Department of Hunan Province", "awards": [{"funding_scheme": ["Key Research Project of Higher Education Institutions"], "award_ids": ["202006"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Partial funding came from the Qinghai Provincial Department of Science and Technology Key Research and Development Program award 2022-ZJ-925. We acknowledge Professor Natalia Petrov of the University of Melbourne for identifying the problem addressed in Theorem 2.1."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Qinghai Provincial Department of Science and Technology", "awards": [{"funding_scheme": ["Key Research and Development Program"], "award_ids": ["2022-ZJ-925"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Satellite and reanalysis datasets were generously provided by the Earth Observation (EO) program of Universität Hannover's the Earth Observation Research Grant, EC, FAU, and NSFC via the Ministry of Agriculture and Rural Affairs. This study received support from a University of Malaya grant R-263-000-532-112 funded by CIHR."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Universität Hannover", "awards": [{"funding_scheme": ["Earth Observation Research Grant"], "award_ids": [], "award_title": []}]}, {"funder_name": "EC", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "FAU", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "NSFC", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Ministry of Agriculture and Rural Affairs", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "University of Malaya", "awards": [{"funding_scheme": [], "award_ids": ["R-263-000-532-112"], "award_title": []}]}, {"funder_name": "CIHR", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Partial support was provided by the Ministry of Higher Education (PG/08/103/26133), Diabetes UK (BITS Pilani), as well as the Environmental Technology Innovation Program (KEITI)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Ministry of Higher Education", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Diabetes UK", "awards": [{"funding_scheme": [], "award_ids": ["PG/08/103/26133"], "award_title": []}]}, {"funder_name": "BITS Pilani", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "KEITI", "awards": [{"funding_scheme": ["Environmental Technology Innovation Program"], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "We express gratitude to K. Patel (LIHD) for sharing the modified strain. Technical assistance from the Cytometry Services Platform CYTOSERV of the Institut Mondor is acknowledged. Support was obtained from the Laboratory of Excellence NEURO-Ex (award 218318); this laboratory receives its funding from the Sandia National Laboratories through the Laboratory Directed Research and Development program, specifically grant MOU 04-S-383. The work of CL, DB, FD, AM, MEN, GN, and LB was supported by funding from Henan University as well as Agency for Administration of University and Research, with additional support from the China Postdoctoral Science Foundation under the \"FRM Group\" mechanism (2019TQ0011)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Sandia National Laboratories", "awards": [{"funding_scheme": ["Laboratory Directed Research and Development"], "award_ids": ["218318", "MOU 04-S-383"], "award_title": []}]}, {"funder_name": "Henan University", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Agency for Administration of University and Research", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "China Postdoctoral Science Foundation", "awards": [{"funding_scheme": [], "award_ids": ["2019TQ0011"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Funding was obtained from MOH via Evaluating Long-term Health Outcomes of Community-Based Chronic Disease Management Programs, as well as the Ministry for Innovation and Technology, as well as MOH (Predictive Modeling of Epidemiological Patterns in Urban Populations for Enhanced Healthcare Resource Allocation, Evaluating Long-term Health Outcomes of Community-Based Chronic Disease Management Programs)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "MOH", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": ["Evaluating Long-term Health Outcomes of Community-Based Chronic Disease Management Programs", "Predictive Modeling of Epidemiological Patterns in Urban Populations for Enhanced Healthcare Resource Allocation"]}]}, {"funder_name": "Ministry for Innovation and Technology", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This work received support from SEDF (21SG42, 21SG42), as well as the Shanghai Municipal Education Commission, as well as Kyung Hee University."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "SEDF", "awards": [{"funding_scheme": [], "award_ids": ["21SG42"], "award_title": []}]}, {"funder_name": "Shanghai Municipal Education Commission", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Kyung Hee University", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "The author thanks Maria Torres, David Chen, Laura Kowalski, and Robert Nielsen for illuminating conversations about their research on the Hamilton-Jacobi framework. Appreciation is also extended to Anna Virtanen for her constructive feedback on this manuscript. This work received partial funding from the Ferdowsi University of Mashhad Fundamental Research Excellence Grant."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Ferdowsi University of Mashhad", "awards": [{"funding_scheme": ["Fundamental Research Excellence Grant"], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "J.D.H. acknowledges the support of the National Science Foundation. N.S.G. acknowledges financial support from the Foundation for Food and Agriculture Research (FFAR) Fellows program. We acknowledge support of the IGI LGR ERA, GlaxoSmithKline, and Citris/Banatao Seed Funding. We acknowledge support of a Burroughs Wellcome Fund Career Award at the Scientific Interface (CASI) (to M.P.L.), a Dreyfus foundation award (to M.P.L.), an NIH MIRA award (to M.P.L.), an NSF CAREER award (to M.P.L), an NSF CGEM award (to M.P.L.), a FFAR Young Investigator award (to M.P.L.), a CZI investigator award (to M.P.L), a Sloan Foundation Award (to M.P.L.), a and Moore Foundation Award (to M.P.L.). M.P.L. is a Chan Zuckerberg Biohub investigator, a Hellen Wills Neuroscience Institute Investigator, and an IGI Investigator. R.L.P. acknowledges support from the Schmidt Science Fellows program, in partnership with the Rhodes Trust. Efforts to apply XSI for studying ssDNA on SWCNTs are supported in part by National Cancer Institute grants Structural Biology of DNA Repair (SBDR) CA092584 and CA220430. XSI data was collected at the Advanced Light Source (ALS) beamline SIBYLS which is supported by the DOE-BER IDAT DE-AC02-05CH11231 and NIGMS ALS-ENABLE (P30 GM124169 and S10OD018483). This work benefited from the use of the SasView application, originally developed under NSF award DMR- 0520547. SasView contains code developed with funding from the European Union's | Horizon 2020 research and innovation program under the SINE2020 project, grant agreement No 654000. We thank the staff at the University of California, Berkeley Electron Microscope Laboratory for advice and assistance in electron microscopy sample preparation and data collection. Additional electron microscopy was conducted at the National Center for Electron Microscopy at the Molecular Foundry, Lawrence Berkeley National Laboratory. Work at the Molecular Foundry was supported by the Office of Science, Office of Basic Energy Sciences, of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231. Molecular graphics and analyses performed with UCSF Chimera, developed by the Resource for Biocomputing, Visualization, and Informatics at the University of California, San Francisco, with support from NIH P41GM103311. We would like to acknowledge the use of medical clipart from BioRender.com. Thank you to Dr. Daniel Murray, Dr. Lee Joon Kim, Dr. Abraham Beyene, Dr. James Holton, Dr. Andrew Crothers, and Dr. Michal Hammel for their helpful feedback and editing of this manuscript. Thank you to Brandon Russel for enabling remote modeling operations. Thank you to Elizabeth Voke for making this long-distance project possible by always being ready to send images of written lab notebooks during the manuscript writing process."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "National Science Foundation", "awards": [{"funding_scheme": [], "award_ids": ["DMR- 0520547"], "award_title": []}]}, {"funder_name": "Foundation for Food and Agriculture Research (FFAR)", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "IGI", "awards": [{"funding_scheme": ["LGA ERA"], "award_ids": [], "award_title": []}]}, {"funder_name": "GlaxoSmithKline", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Citris/Banatao", "awards": [{"funding_scheme": ["Seed Funding"], "award_ids": [], "award_title": []}]}, {"funder_name": "Burroughs Wellcome Fund", "awards": [{"funding_scheme": ["Career Award"], "award_ids": [], "award_title": []}]}, {"funder_name": "Dreyfus foundation", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "NIH", "awards": [{"funding_scheme": ["MIRA"], "award_ids": ["P41GM103311"], "award_title": []}]}, {"funder_name": "NSF", "awards": [{"funding_scheme": ["CAREER"], "award_ids": [], "award_title": []}, {"funding_scheme": ["CGEM"], "award_ids": [], "award_title": []}]}, {"funder_name": "FFAR", "awards": [{"funding_scheme": ["Young Investigator award"], "award_ids": [], "award_title": []}]}, {"funder_name": "CZI", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Sloan Foundation", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Moore Foundation", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Schmidt Science", "awards": [{"funding_scheme": ["Fellows program"], "award_ids": [], "award_title": []}]}, {"funder_name": "Rhodes Trust", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "National Cancer Institute", "awards": [{"funding_scheme": [], "award_ids": ["CA092584", "CA220430"], "award_title": ["Structural Biology of DNA Repair"]}]}, {"funder_name": "DOE-BER", "awards": [{"funding_scheme": [], "award_ids": ["IDAT DE-AC02-05CH11231"], "award_title": ["SasView"]}]}, {"funder_name": "NIGMS", "awards": [{"funding_scheme": [], "award_ids": ["P30 GM124169", "S10OD018483"], "award_title": ["ALS-ENABLE"]}]}, {"funder_name": "European Union", "awards": [{"funding_scheme": ["Horizon 2020 research and innovation program"], "award_ids": ["654000"], "award_title": ["SINE2020 project"]}]}, {"funder_name": "Office of Science, Office of Basic Energy Sciences, of the U.S. Department of Energy", "awards": [{"funding_scheme": [], "award_ids": ["DE-AC02-05CH11231"], "award_title": []}]}]} | {"doi": "10.1101/2023.05.04.539504"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "This work was supported by the U.S. Department of Energy, Energy Efficiency & Renewable Energy (EERE) under Award No. DE-EE0010432."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "U.S. Department of Energy, Energy Efficiency & Renewable Energy (EERE)", "awards": [{"funding_scheme": [], "award_ids": ["DE-EE0010432"], "award_title": []}]}]} | {"doi": "10.26434/chemrxiv-2025-1z9j7"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "We gratefully acknowledge NIST. Support was received under grant 70NANB19H059."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "NIST", "awards": [{"funding_scheme": [], "award_ids": ["70NANB19H059"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Funding for this study was provided by Agencia Nacional de Investigación y Desarrollo and the National Polytechnic Institute, whose support we gratefully acknowledge. A Victoria University of Wellington Postgraduate Research Excellence Award via the Postgraduate Doctoral Scholarship (corporate partner: Global Marine Trust) supported JH, who also obtained additional financial assistance from the Cambridge Commonwealth European and International Trust and the Flemish Government. The American University of Beirut funded field activities in Seychelles. RB's research was financially supported through Agencia Nacional de Investigación y Desarrollo and the Pathfinder Challenge of the EIC. The field work and sample collection efforts of BRT were generously funded by Agencia Nacional de Investigación y Desarrollo and the Coastal Ecosystem Protection Fund (CEPF)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Agencia Nacional de Investigación y Desarrollo", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "National Polytechnic Institute", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Victoria University of Wellington", "awards": [{"funding_scheme": ["Postgraduate Research Excellence Award", "Postgraduate Doctoral Scholarship"], "award_ids": [], "award_title": []}]}, {"funder_name": "Cambridge Commonwealth European and International Trust", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Flemish Government", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "American University of Beirut", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "EIC", "awards": [{"funding_scheme": ["Pathfinder Challenge"], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Acknowledgments. The authors acknowledge support of this research by Ministry of Innovative Development of the Republic of Uzbekistan (Grant No. FZ-20200929224)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Ministry of Innovative Development of the Republic of Uzbekistan", "awards": [{"funding_scheme": [], "award_ids": ["FZ-20200929224"], "award_title": []}]}]} | {"doi": "10.48550/arxiv.2304.11610"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "We thank Hsuan-Chi Liu, Chin-Kun Fu, and Shih-Yao Mao for helping with participant recruitment and experiment preparation. We also thank Felix B. Kern and Amit Yaron for proofreading. This study was supported by Japan Society for the Promotion of Science, Japan (to Y.T.H), Ministry of Science and Technology of Taiwan, Taiwan (MOST 106-2420-H-002 -008-MY2 and MOST 109-2410-H-002-106-MY3) (to C.W), and World Premier International Research Center Initiative (WPI), MEXT, Japan (to Z.C.C.)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Japan Society for the Promotion of Science, Japan", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Ministry of Science and Technology of Taiwan, Taiwan", "awards": [{"funding_scheme": ["MOST"], "award_ids": ["MOST 106-2420-H-002 -008-MY2", "MOST 109-2410-H-002-106-MY3"], "award_title": []}]}, {"funder_name": "World Premier International Research Center Initiative (WPI)", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "MEXT, Japan", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {"doi": "10.1101/2023.11.20.567982"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "J.B. received funding from Patheon Viral Vector Services to undertake this research. | The authors greatly acknowledge funding and support from Patheon, and greatly thank Anne Chen for her assistance with project management."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "Patheon Viral Vector Services", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {"doi": "10.26434/chemrxiv-2023-ppzsj"} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Our research is supported by the Collaborative Research Grant award 110121, which obtains resources from the Chemical Industry Research Grant of Der Fonds der Chemischen Industrie. Fédération Wallonie-Bruxelles contributes additional partial support."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": null, "awards": [{"funding_scheme": ["Collaborative Research Grant"], "award_ids": [], "award_title": []}]}, {"funder_name": "Der Fonds der Chemischen Industrie", "awards": [{"funding_scheme": ["Chemical Industry Research Grant"], "award_ids": ["110121"], "award_title": []}]}, {"funder_name": "Fédération Wallonie-Bruxelles", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Financial support was provided to K.R.N. by the National Office of Philosophy and Social Sciences. Both T.L.W. and P.D.M. received funding from King Khalid University for neurological disease research."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "National Office of Philosophy and Social Sciences", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "King Khalid University", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "Funding agencies supporting this project include the United States National Library of Medicine (award NHLBI K01HL153942), United States Army Research Laboratory (award W911NF-21-2-0127,W911NF-22-2-0003), South African College (award 256494), and the Advanced Research Collaborative Grant (award NA140436)."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "United States National Library of Medicine", "awards": [{"funding_scheme": [], "award_ids": ["NHLBI K01HL153942"], "award_title": []}]}, {"funder_name": "United States Army Research Laboratory", "awards": [{"funding_scheme": [], "award_ids": ["W911NF-21-2-0127,W911NF-22-2-0003"], "award_title": []}]}, {"funder_name": "South African College", "awards": [{"funding_scheme": [], "award_ids": ["256494"], "award_title": []}]}, {"funder_name": null, "awards": [{"funding_scheme": ["Advanced Research Collaborative Grant"], "award_ids": ["NA140436"], "award_title": []}]}]} | {} |
[{"role": "system", "content": "Your task is to extract all funding entities and their associated grant or award identifiers from a single funding statement. You must carefully identify each funder and their grant IDs.\n\nReturn the result as a structured list of objects. Each object in the list should represent a distinct funder and contain the following fields:\n- `funder_name`: The name of the funding organization or program.\n- `award_ids`: A list of all grant numbers or identifiers associated with the funder.\n\n### Detailed Instructions and Rules:\n\n1. **Funder Identification (`funder_name`):**\n * **Use Literal Names:** Extract the funder's name exactly as it appears in the text. Do not use external knowledge to expand acronyms. For example, if the text says \"NSF\", the `funder_name` is \"NSF\", not \"National Science Foundation\" (unless the full name is also provided in the text).\n * **Capture Full Hierarchical Names:** If a funder is described with a specific sub-division, department, or office (e.g., \"US Department of Energy, Fusion Energy Sciences\"), capture this entire, more specific name as the `funder_name`.\n * **Programs and Schemes:** If a funding program or scheme is named alongside a clear parent organization, ignore the program name and use only the organization as the `funder_name`. If no parent organization is clearly identified, use the program name as the `funder_name`.\n * **Identify All Involved Parties:** In complex statements, multiple entities may be considered funders.\n * For phrases like \"...carried out within the framework of the **EUROfusion Consortium**, funded by the **European Union**...\", both \"EUROfusion Consortium\" and \"European Union\" should be extracted as separate funders.\n * Extract organizations mentioned in related disclaimers, even if they are being absolved of responsibility (e.g., from \"...Neither the European Union nor the **European Commission** can be held responsible...\", extract \"European Commission\" as a funder).\n * **Consolidate Duplicates:** If the exact same funder is mentioned multiple times, create only one object for it and aggregate all its associated `award_ids`.\n\n2. **Award IDs (`award_ids`):**\n * Extract all alphanumeric codes that represent grant numbers, contract numbers, or project identifiers.\n * Accurately associate each ID with its correct funder. For example, in \"...EUROfusion Consortium... (Grant Agreement No 101052200...)\", \"101052200\" should be associated with \"EUROfusion Consortium\".\n * A single funder can have multiple `award_ids`. Collect them all into a single list for that funder.\n * If a funder has no associated grant numbers, use an empty list.\n\n3. **General Approach:**\n * Parse the entire statement carefully, paying close attention to prepositions like \"by\", \"via\", and \"within\".\n * Create one object per unique funder identified.\n * Aggregate all awards for a single funder into its corresponding `award_ids` list.\n * Do not extract acknowledgements to individuals as funders.\n\n### Examples:\n\n**Example 1 (Simple):**\n\nInput: \"This material is based upon work supported by the National Science Foundation under Grant No. 1116589.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"1116589\"]\n }\n]\n```\n\n**Example 2 (Multiple funders, program as funder):**\n\nInput: \"This work was supported by grants from the National Natural Science Foundation of China (no. 81772706, no. 81802525, and no.821172817) and the National Key Research and Development Project (no. 2019YFC1316005).\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Natural Science Foundation of China\",\n \"award_ids\": [\"81772706\", \"81802525\", \"821172817\"]\n },\n {\n \"funder_name\": \"National Key Research and Development Project\",\n \"award_ids\": [\"2019YFC1316005\"]\n }\n]\n```\n\n**Example 3 (Many funders, hierarchical names):**\n\nInput: \"This work was supported by NSF ABI 1564659, NSF CAREER 2042516 to AS. This work was funded by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Hatch Program under accession number 1008480 and funds from the University of Kentucky Bobby C. Pass Research Professorship to JJO. This research was supported in part by a Research Support Grant from the University of Kentucky Office of the Vice President for Research to DWW and JJO. This research includes calculations carried out on HPC resources supported in part by the National Science Foundation through major research instrumentation grant number 1625061 and by the US Army Research Laboratory under contract number W911NF- 16-2- 0189.\"\n\nOutput:\n```json\n[\n {\n \"funder_name\": \"National Science Foundation\",\n \"award_ids\": [\"ABI 1564659\", \"CAREER 2042516\", \"1625061\"]\n },\n {\n \"funder_name\": \"National Institute of Food and Agriculture, U.S. Department of Agriculture\",\n \"award_ids\": [\"1008480\"]\n },\n {\n \"funder_name\": \"University of Kentucky Bobby C. Pass Research Professorship\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"University of Kentucky Office of the Vice President for Research\",\n \"award_ids\": []\n },\n {\n \"funder_name\": \"US Army Research Laboratory\",\n \"award_ids\": [\"W911NF- 16-2- 0189\"]\n }\n]\n```"}, {"role": "user", "content": "ELW acknowledges the hospitality of the IAS during the writing of this paper. Astrophysics research at the IAS receives support from University of Milano-Bicocca award 2015-ATE-0447 and the Colorado State College of Agriculture and Mechanic Arts. Peking University datasets were developed by the Korea Health Industry Development Institute under the guidance of the Université de Montréal and were provided by the University of São Paulo publication makes use of data products from the Ludwig Maximilian University which is a joint project of the University of Massachusetts and the École Polytechnique and University of Amsterdam funded by the MHESI and the University of Milano-Bicocca."}] | funding_extraction | {"method": "rule", "ground_truth": [{"funder_name": "University of Milano-Bicocca", "awards": [{"funding_scheme": [], "award_ids": ["2015-ATE-0447"], "award_title": []}]}, {"funder_name": "Colorado State College of Agriculture and Mechanic Arts", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "Korea Health Industry Development Institute", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}, {"funder_name": "MHESI", "awards": [{"funding_scheme": [], "award_ids": [], "award_title": []}]}]} | {} |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.