bourdoiscatie commited on
Commit
1536dfe
1 Parent(s): c9bc1c7

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +97 -0
README.md ADDED
@@ -0,0 +1,97 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - fr
4
+ license:
5
+ - unknown
6
+ size_categories:
7
+ - 10K<n<100K
8
+ task_categories:
9
+ - token-classification
10
+ tags:
11
+ - ner
12
+ ---
13
+
14
+ # wikiner_fr_prompt_ner
15
+ ## Summary
16
+
17
+ **wikiner_fr_prompt_ner** is a subset of the [**Dataset of French Prompts (DFP)**]().
18
+ It contains **X** rows that can be used for a name entity recognition task.
19
+ The original data (without prompts) comes from the dataset [wikiner](https://huggingface.co/datasets/Jean-Baptiste/wikiner_fr) by Nothman et al. where only the French part has been kept.
20
+ 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](https://huggingface.co/datasets/bigscience/xP3) dataset by Muennighoff et al.
21
+
22
+
23
+ ## Prompts used
24
+ ### List
25
+ 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.
26
+
27
+ ```
28
+ 'Extraire les classes des mots du texte suivant : '+text,
29
+ 'Extrais les entitées nommées du texte suivant : '+text,
30
+ 'Extrayez les entitées nommées du texte suivant : '+text,
31
+ 'Isoler les entitées nommées du texte suivant : '+text,
32
+ 'Isole les entitées nommées du texte suivant : '+text,
33
+ 'Isolez les entitées nommées du texte suivant : '+text,
34
+ 'Dégager des entitées nommées dans le texte : '+text,
35
+ 'Dégage des entitées nommées dans le texte : '+text,
36
+ 'Dégagez des entitées nommées dans le texte : '+text,
37
+ 'Générer des entitées nommées issues du texte suivant : '+text,
38
+ 'Génère des entitées nommées issues du texte suivant : '+text,
39
+ 'Générez des entitées nommées issues du texte suivant : '+text,
40
+ 'Trouver les entitées nommées du texte : '+text,
41
+ 'Trouve les entitées nommées du texte : '+text,
42
+ 'Trouvez les entitées nommées du texte : '+text,
43
+ 'Repérer les entitées nommées présentes dans le texte suivant : '+text,
44
+ 'Repère les entitées nommées présentes dans le texte suivant : '+text,
45
+ 'Repérez les entitées nommées présentes dans le texte suivant : '+text,
46
+ 'Indiquer les entitées nommées du texte :'+text,
47
+ 'Indique les entitées nommées du texte : '+text,
48
+ 'Indiquez les entitées nommées du texte : '+text
49
+ ```
50
+
51
+ ### Features used in the prompts
52
+ In the prompt list above, `text` and `targets` have been constructed from:
53
+ ```
54
+ wikiner_fr = load_dataset('Jean-Baptiste/wikiner_fr')
55
+ wikiner_fr['train']['tokens'] = list(map(lambda i: ' '.join(wikiner_fr['train']['tokens'][i]), range(len(wikiner_fr['train']['tokens']))))
56
+ wikiner_fr['train']['ner_tags'] = list(map(lambda x: x.replace("[","").replace("]","").replace('0','O').replace('1','LOC').replace('2','PER').replace('3','O').replace('4','ORG'), map(str, wikiner_fr['train']['ner_tags'])))
57
+ ```
58
+
59
+
60
+
61
+ # Splits
62
+ - train with X samples
63
+ - dev with Y samples
64
+ - test with Z samples
65
+
66
+
67
+ # How to use?
68
+ ```
69
+ from datasets import load_dataset
70
+ dataset = load_dataset("CATIE-AQ/wikiner_fr_prompt_ner")
71
+ ```
72
+
73
+ # Citation
74
+ ## Original data
75
+ > @article{NOTHMAN2013151,
76
+ title = {Learning multilingual named entity recognition from Wikipedia},
77
+ journal = {Artificial Intelligence},
78
+ volume = {194},
79
+ pages = {151-175},
80
+ year = {2013},
81
+ note = {Artificial Intelligence, Wikipedia and Semi-Structured Resources},
82
+ issn = {0004-3702},
83
+ doi = {https://doi.org/10.1016/j.artint.2012.03.006},
84
+ url = {https://www.sciencedirect.com/science/article/pii/S0004370212000276},
85
+ author = {Joel Nothman and Nicky Ringland and Will Radford and Tara Murphy and James R. Curran},
86
+ keywords = {Named entity recognition, Information extraction, Wikipedia, Semi-structured resources, Annotated corpora, Semi-supervised learning},
87
+ abstract = {We automatically create enormous, free and multilingual silver-standard training annotations for named entity recognition (ner) by exploiting the text and structure of Wikipedia. Most ner systems rely on statistical models of annotated data to identify and classify names of people, locations and organisations in text. This dependence on expensive annotation is the knowledge bottleneck our work overcomes. We first classify each Wikipedia article into named entity (ne) types, training and evaluating on 7200 manually-labelled Wikipedia articles across nine languages. Our cross-lingual approach achieves up to 95% accuracy. We transform the links between articles into ne annotations by projecting the target articleʼs classifications onto the anchor text. This approach yields reasonable annotations, but does not immediately compete with existing gold-standard data. By inferring additional links and heuristically tweaking the Wikipedia corpora, we better align our automatic annotations to gold standards. We annotate millions of words in nine languages, evaluating English, German, Spanish, Dutch and Russian Wikipedia-trained models against conll shared task data and other gold-standard corpora. Our approach outperforms other approaches to automatic ne annotation (Richman and Schone, 2008 [61], Mika et al., 2008 [46]) competes with gold-standard training when tested on an evaluation corpus from a different source; and performs 10% better than newswire-trained models on manually-annotated Wikipedia text.}
88
+ }
89
+
90
+
91
+
92
+ ## This Dataset
93
+
94
+
95
+
96
+ ## License
97
+ Unknow