README.md

---
annotations_creators:
- machine-generated
language:
- en
- fr
- pt
- de
- fr
- it
- es
language_creators:
- found
license:
- mit
multilinguality:
- multilingual
pretty_name: Professor HeidelTime
size_categories:
- 100K<n<1M
source_datasets:
- original
tags:
- Timex
- Timexs
- Temporal Expression
- Temporal Expressions
- Temporal Information
- Timex Identification
- Timex Classification
- Timex Extraction
task_categories:
- token-classification
task_ids:
- parsing
- part-of-speech
- named-entity-recognition
configs:
- config_name: portuguese
  data_files: "portuguese.json"
- config_name: english
  data_files: "english.json"
- config_name: french
  data_files: "french.json"
- config_name: italian
  data_files: "italian.json"
- config_name: spanish
  data_files: "spanish.json"
- config_name: german
  data_files: "german.json"
---
# Professor HeidelTime

[Paper](https://dl.acm.org/doi/10.1145/3583780.3615130)    [GitHub](https://github.com/hmosousa/professor_heideltime)

Professor HeidelTime is a project to create a multilingual corpus weakly labeled with [HeidelTime](https://github.com/HeidelTime/heideltime), a temporal tagger.

## Corpus Details

The weak labeling was performed in six languages. Here are the specifics of the corpus for each language:

| Dataset                 | Language | Documents | From       | To         | Tokens     | Timexs    |
| ----------------------- | -------- | --------- | ---------- | ---------- | ---------- | --------  |
| All the News 2.0        | EN       | 24,642    | 2016-01-01 | 2020-04-02 | 18,755,616 | 254,803   |
| Italian Crime News      | IT       | 9,619     | 2011-01-01 | 2021-12-31 | 3,296,898  | 58,823    |
| German News Dataset     | DE       | 33,266    | 2003-01-01 | 2022-12-31 | 21,617,888 | 348,011   |
| ElMundo News            | ES       | 19,095    | 2005-12-02 | 2021-10-18 | 12,515,410 | 194,043   |
| French Financial News   | FR       | 24,293    | 2017-10-19 | 2021-03-19 | 1,673,053  | 83,431    |
| Público News            | PT       | 27,154    | 2000-11-14 | 2002-03-20 | 5,929,377  | 111,810   |

## Contact

For more information, reach out to [Hugo Sousa](https://hugosousa.net) at <hugo.o.sousa@inesctec.pt>.

This framework is a part of the [Text2Story](https://text2story.inesctec.pt) project. This project is financed by the ERDF – European Regional Development Fund through the North Portugal Regional Operational Programme (NORTE 2020), under the PORTUGAL 2020 and by National Funds through the Portuguese funding agency, FCT - Fundação para a Ciência e a Tecnologia within project PTDC/CCI-COM/31857/2017 (NORTE-01-0145-FEDER-03185).

## Cite

If you use this work, please cite the following [paper](https://dl.acm.org/doi/10.1145/3583780.3615130):

```bibtex
@inproceedings{10.1145/3583780.3615130,
    author = {Sousa, Hugo and Campos, Ricardo and Jorge, Al\'{\i}pio},
    title = {TEI2GO: A Multilingual Approach for Fast Temporal Expression Identification},
    year = {2023},
    isbn = {9798400701245},
    publisher = {Association for Computing Machinery},
    url = {https://doi.org/10.1145/3583780.3615130},
    doi = {10.1145/3583780.3615130},
    booktitle = {Proceedings of the 32nd ACM International Conference on Information and Knowledge Management},
    pages = {5401–5406},
    numpages = {6},
    keywords = {temporal expression identification, multilingual corpus, weak label},
    location = {Birmingham, United Kingdom},
    series = {CIKM '23}
}
```