README.md

metadata

license: cc0-1.0
task_categories:
  - text-generation
  - token-classification
language:
  - en
tags:
  - funding
  - research-funding
  - scholarly-metadata
  - preprints
  - arxiv
size_categories:
  - 1K<n<10K
configs:
  - config_name: default
    data_files:
      - split: train
        path: train.jsonl
      - split: test
        path: test.jsonl

Preprint Funding Metadata

A dataset of funding statements extracted from arXiv preprints with structured funder and award annotations.

Dataset Description

This dataset contains funding statements from scientific preprints along with manually annotated structured metadata identifying funders and their associated awards.

Data Fields

• doi: The DOI of the preprint (e.g., "10.48550/arxiv.2303.07677")
• funding_statement: The raw funding acknowledgment text extracted from the paper
• funders: Array of funder objects containing:
  • funder_name: Name of the funding organization (may be null if only a program is mentioned)
  • awards: Array of award objects containing:
    • funding_scheme: Array of funding program/scheme names
    • award_ids: Array of grant/award identifiers
    • award_title: Array of award titles (if available)

Dataset Statistics

Split	Examples
Train	1,388
Test	347

Usage

from datasets import load_dataset

dataset = load_dataset("cometadata/preprint-funding")

# Access train split
train_data = dataset["train"]

# Access test split
test_data = dataset["test"]

Example

{
  "doi": "10.48550/arxiv.2303.07677",
  "funding_statement": "This work was supported in part by the Key R&D Program of Zhejiang under Grant 2022C01018, and by the National Natural Science Foundation of China under Grants U21B2001 and 61973273.",
  "funders": [
    {
      "funder_name": null,
      "awards": [
        {
          "funding_scheme": ["Key R&D Program of Zhejiang"],
          "award_ids": ["2022C01018"],
          "award_title": []
        }
      ]
    },
    {
      "funder_name": "National Natural Science Foundation of China",
      "awards": [
        {
          "funding_scheme": [],
          "award_ids": ["U21B2001", "61973273"],
          "award_title": []
        }
      ]
    }
  ]
}

License

This dataset is released under CC0 (Public Domain).