Datasets:
Tasks:
Token Classification
Modalities:
Text
Formats:
parquet
Languages:
French
Size:
10K - 100K
ArXiv:
License:
Update README.md
Browse files
README.md
CHANGED
@@ -28,15 +28,53 @@ configs:
|
|
28 |
|
29 |
This dataset is a manually revised version of 20% of the French proportion of [WikiNER](https://doi.org/10.1016/j.artint.2012.03.006).
|
30 |
The original dataset is currently available [here](https://figshare.com/articles/dataset/Learning_multilingual_named_entity_recognition_from_Wikipedia/5462500), based on which WikiNER-fr-gold is created.
|
31 |
-
|
32 |
-
The dataset is formatted in CoNLL style with three columns: text, POS and NER.
|
33 |
The entities are annotated using the BIOES scheme.
|
34 |
The POS tags are not revised i.e. remain the same as the original dataset.
|
35 |
|
36 |
For more information on the revision details, please refer to our paper [WikiNER-fr-gold: A Gold-Standard NER Corpus](https://arxiv.org/abs/2411.00030).
|
37 |
|
38 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
39 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
40 |
|
41 |
```
|
42 |
@misc{cao2024wikinerfrgoldgoldstandardnercorpus,
|
|
|
28 |
|
29 |
This dataset is a manually revised version of 20% of the French proportion of [WikiNER](https://doi.org/10.1016/j.artint.2012.03.006).
|
30 |
The original dataset is currently available [here](https://figshare.com/articles/dataset/Learning_multilingual_named_entity_recognition_from_Wikipedia/5462500), based on which WikiNER-fr-gold is created.
|
|
|
|
|
31 |
The entities are annotated using the BIOES scheme.
|
32 |
The POS tags are not revised i.e. remain the same as the original dataset.
|
33 |
|
34 |
For more information on the revision details, please refer to our paper [WikiNER-fr-gold: A Gold-Standard NER Corpus](https://arxiv.org/abs/2411.00030).
|
35 |
|
36 |
+
The dataset is available in two formats.
|
37 |
+
The CoNLL version contains three columns: text, POS and NER.
|
38 |
+
The JSON version is downloadable using the `datasets` library.
|
39 |
+
Originally conceived as a test set, there is no recommended train/dev/test split.
|
40 |
+
The downloaded dataset is by default labeled `train`.
|
41 |
+
|
42 |
+
```python
|
43 |
+
from datasets import load_dataset
|
44 |
+
|
45 |
+
ds = load_dataset('danrun/WikiNER-fr-gold')
|
46 |
+
|
47 |
+
ds['train'][0]
|
48 |
+
# {'ner_tags': [...], 'tokens': [...], 'pos_tags': [...]}
|
49 |
+
```
|
50 |
+
|
51 |
+
The NER tags are indexed using the following table (see `labels.json`):
|
52 |
|
53 |
+
```
|
54 |
+
{
|
55 |
+
'O': 0,
|
56 |
+
'B-PER': 1,
|
57 |
+
'I-PER': 2,
|
58 |
+
'E-PER': 3,
|
59 |
+
'S-PER': 4,
|
60 |
+
'B-LOC': 5,
|
61 |
+
'I-LOC': 6,
|
62 |
+
'E-LOC': 7,
|
63 |
+
'S-LOC': 8,
|
64 |
+
'B-ORG': 9,
|
65 |
+
'I-ORG': 10,
|
66 |
+
'E-ORG': 11,
|
67 |
+
'S-ORG': 12,
|
68 |
+
'B-MISC': 13,
|
69 |
+
'I-MISC': 14,
|
70 |
+
'E-MISC': 15,
|
71 |
+
'S-MISC': 16
|
72 |
+
}
|
73 |
+
```
|
74 |
+
|
75 |
+
|
76 |
+
|
77 |
+
## Citation
|
78 |
|
79 |
```
|
80 |
@misc{cao2024wikinerfrgoldgoldstandardnercorpus,
|