bourdoiscatie's picture
Update README.md
64c99bf
metadata
language:
  - fr
license:
  - unknown
size_categories:
  - 100K<n<1M
task_categories:
  - token-classification
tags:
  - ner
  - DFP
  - french prompts
annotations_creators:
  - found
language_creators:
  - found
multilinguality:
  - monolingual
source_datasets:
  - wikiann

wikiann_fr_prompt_ner

Summary

wikiann_fr_prompt_ner is a subset of the Dataset of French Prompts (DFP).
It contains 840,000 rows that can be used for a name entity recognition task.
The original data (without prompts) comes from the dataset wikiann by Pan et al. where only the French part has been kept.
A list of prompts (see below) was then applied in order to build the input and target columns and thus obtain the same format as the xP3 dataset by Muennighoff et al.

Prompts used

List

21 prompts were created for this dataset. The logic applied consists in proposing prompts in the indicative tense, in the form of tutoiement and in the form of vouvoiement.

'Extraire les entités nommées du texte suivant : '+text,
'Extrais les entités nommées du texte suivant : '+text,
'Extrayez les entités nommées du texte suivant : '+text,
'Isoler les entités nommées du texte suivant : '+text,
'Isole les entités nommées du texte suivant : '+text,
'Isolez les entités nommées du texte suivant : '+text,
'Dégager des entités nommées dans le texte : '+text,
'Dégage des entités nommées dans le texte : '+text,
'Dégagez des entités nommées dans le texte : '+text,
'Générer des entités nommées issues du texte suivant : '+text,
'Génère des entités nommées issues du texte suivant : '+text,
'Générez des entités nommées issues du texte suivant : '+text,
'Trouver les entités nommées du texte : '+text,
'Trouve les entités nommées du texte : '+text,
'Trouvez les entités nommées du texte : '+text,
'Repérer les entités nommées présentes dans le texte suivant : '+text,
'Repère les entités nommées présentes dans le texte suivant : '+text,
'Repérez les entités nommées présentes dans le texte suivant : '+text,
'Indiquer les entités nommées du texte :'+text,
'Indique les entités nommées du texte : '+text,
'Indiquez les entités nommées du texte : '+text

Features used in the prompts

In the prompt list above, text and targets have been constructed from:

wikiann = load_dataset('tner/wikiann','fr') 
wikiann['train']['tokens'] = list(map(lambda i: ' '.join(wikiann['train']['tokens'][i]), range(len(wikiann['train']['tokens']))))
wikiann['train']['tags'] = list(map(lambda x: x.replace("[","").replace("]","").replace('0','B-LOC').replace('1','B-ORG').replace('2','B-PER').replace('3','I-LOC').replace('4','I-ORG').replace('5','I-PER').replace('6','O'), map(str, wikiann['train']['tags'])))

Splits

  • train with 420,000 samples
  • valid with 210,000 samples
  • test with 210,000 samples

How to use?

from datasets import load_dataset
dataset = load_dataset("CATIE-AQ/wikiann_fr_prompt_ner")

Citation

Original data

@inproceedings{pan-etal-2017-cross, title = "Cross-lingual Name Tagging and Linking for 282 Languages", author = "Pan, Xiaoman and Zhang, Boliang and May, Jonathan and Nothman, Joel and Knight, Kevin and Ji, Heng", booktitle = "Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)", month = jul, year = "2017", address = "Vancouver, Canada", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/P17-1178", doi = "10.18653/v1/P17-1178", pages = "1946--1958" }

This Dataset

@misc {centre_aquitain_des_technologies_de_l'information_et_electroniques_2023,
author = { {Centre Aquitain des Technologies de l'Information et Electroniques} },
title = { DFP (Revision 1d24c09) },
year = 2023,
url = { https://huggingface.co/datasets/CATIE-AQ/DFP },
doi = { 10.57967/hf/1200 },
publisher = { Hugging Face }
}

License

Unknow