v1.0.0
Browse files- README.md +89 -0
- RuNNE.py +94 -0
- data/dev.jsonl +0 -0
- data/test.jsonl +0 -0
- data/train.jsonl +0 -0
- ent_types.txt +29 -0
README.md
ADDED
@@ -0,0 +1,89 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
languages:
|
3 |
+
- ru
|
4 |
+
multilinguality:
|
5 |
+
- monolingual
|
6 |
+
pretty_name: RuNNE
|
7 |
+
task_categories:
|
8 |
+
- structure-prediction
|
9 |
+
task_ids:
|
10 |
+
- named-entity-recognition
|
11 |
+
---
|
12 |
+
|
13 |
+
# RuNNE dataset
|
14 |
+
|
15 |
+
## Table of Contents
|
16 |
+
- [Dataset Description](#dataset-description)
|
17 |
+
- [Dataset Structure](#dataset-structure)
|
18 |
+
- [Citation Information](#citation-information)
|
19 |
+
- [Contacts](#contacts)
|
20 |
+
|
21 |
+
## Dataset Description
|
22 |
+
Part of NEREL dataset (https://arxiv.org/abs/2108.13112), a Russian dataset
|
23 |
+
for named entity recognition and relation extraction, used in RuNNE (2022)
|
24 |
+
competition (https://github.com/dialogue-evaluation/RuNNE).
|
25 |
+
|
26 |
+
Entities may be nested (see https://arxiv.org/abs/2108.13112).
|
27 |
+
|
28 |
+
Entity types list:
|
29 |
+
* AGE
|
30 |
+
* AWARD
|
31 |
+
* CITY
|
32 |
+
* COUNTRY
|
33 |
+
* CRIME
|
34 |
+
* DATE
|
35 |
+
* DISEASE
|
36 |
+
* DISTRICT
|
37 |
+
* EVENT
|
38 |
+
* FACILITY
|
39 |
+
* FAMILY
|
40 |
+
* IDEOLOGY
|
41 |
+
* LANGUAGE
|
42 |
+
* LAW
|
43 |
+
* LOCATION
|
44 |
+
* MONEY
|
45 |
+
* NATIONALITY
|
46 |
+
* NUMBER
|
47 |
+
* ORDINAL
|
48 |
+
* ORGANIZATION
|
49 |
+
* PENALTY
|
50 |
+
* PERCENT
|
51 |
+
* PERSON
|
52 |
+
* PRODUCT
|
53 |
+
* PROFESSION
|
54 |
+
* RELIGION
|
55 |
+
* STATE_OR_PROVINCE
|
56 |
+
* TIME
|
57 |
+
* WORK_OF_ART
|
58 |
+
|
59 |
+
## Dataset Structure
|
60 |
+
There are two "configs" or "subsets" of the dataset.
|
61 |
+
Using
|
62 |
+
`load_dataset('MalakhovIlya/RuNNE', 'ent_types')['ent_types']`
|
63 |
+
you can download list of entity types (
|
64 |
+
Dataset({
|
65 |
+
features: ['type'],
|
66 |
+
num_rows: 29
|
67 |
+
})
|
68 |
+
)
|
69 |
+
Using
|
70 |
+
`load_dataset('MalakhovIlya/RuNNE', 'data')` or `load_dataset('MalakhovIlya/RuNNE')`
|
71 |
+
you can download the data itself (DatasetDict)
|
72 |
+
|
73 |
+
Dataset consists of 3 splits: "train", "test" and "dev". Each of them contains text document. "Train" and "test" splits also contain annotated entities, "dev" doesn't.
|
74 |
+
Each entity is represented by a string of the following format: "\<start> \<stop> \<type>", where \<start> is a position of the first symbol of entity in text, \<stop> is the last symbol position in text and \<type> is a one of the aforementioned list of types.
|
75 |
+
|
76 |
+
P.S.
|
77 |
+
Original NEREL dataset also contains relations, events and linked entities, but they were not added here yet ¯\\\_(ツ)_/¯
|
78 |
+
|
79 |
+
## Citation Information
|
80 |
+
@article{Artemova2022runne,
|
81 |
+
title={{RuNNE-2022 Shared Task: Recognizing Nested Named Entities}},
|
82 |
+
author={Artemova, Ekaterina and Zmeev, Maksim and Loukachevitch, Natalia and Rozhkov, Igor and Batura, Tatiana and Braslavski, Pavel and Ivanov, Vladimir and Tutubalina, Elena},
|
83 |
+
journal={Computational Linguistics and Intellectual Technologies: Proceedings of the International Conference "Dialog"},
|
84 |
+
year={2022}
|
85 |
+
}
|
86 |
+
|
87 |
+
## Contacts
|
88 |
+
Malakhov Ilya
|
89 |
+
Telegram - https://t.me/noname_4710
|
RuNNE.py
ADDED
@@ -0,0 +1,94 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import datasets
|
2 |
+
import json
|
3 |
+
|
4 |
+
_NAME = 'RuNNE'
|
5 |
+
_CITATION = '''
|
6 |
+
@article{Artemova2022runne,
|
7 |
+
title={{RuNNE-2022 Shared Task: Recognizing Nested Named Entities}},
|
8 |
+
author={Artemova, Ekaterina and Zmeev, Maksim and Loukachevitch,
|
9 |
+
Natalia and Rozhkov, Igor and Batura, Tatiana and Braslavski,
|
10 |
+
Pavel and Ivanov, Vladimir and Tutubalina, Elena},
|
11 |
+
journal={Computational Linguistics and Intellectual Technologies:
|
12 |
+
Proceedings of the International Conference "Dialog"},
|
13 |
+
year={2022}
|
14 |
+
}
|
15 |
+
'''.strip()
|
16 |
+
_DESCRIPTION = 'A Russian Dataset with Nested Named Entities'
|
17 |
+
_HOMEPAGE = 'https://github.com/dialogue-evaluation/RuNNE'
|
18 |
+
_VERSION = '1.0.0'
|
19 |
+
|
20 |
+
|
21 |
+
class RuNNEBuilder(datasets.GeneratorBasedBuilder):
|
22 |
+
_DATA_URLS = {
|
23 |
+
'train': 'data/train.jsonl',
|
24 |
+
'test': 'data/test.jsonl',
|
25 |
+
'dev': 'data/dev.jsonl'
|
26 |
+
}
|
27 |
+
_ENTITY_TYPES_URLS = {
|
28 |
+
'ent_types': 'ent_types.txt'
|
29 |
+
}
|
30 |
+
VERSION = datasets.Version(_VERSION)
|
31 |
+
BUILDER_CONFIGS = [
|
32 |
+
datasets.BuilderConfig('data',
|
33 |
+
version=VERSION,
|
34 |
+
description='Data'),
|
35 |
+
datasets.BuilderConfig('ent_types',
|
36 |
+
version=VERSION,
|
37 |
+
description='Entity types list')
|
38 |
+
]
|
39 |
+
DEFAULT_CONFIG_NAME = 'data'
|
40 |
+
|
41 |
+
def _info(self) -> datasets.DatasetInfo:
|
42 |
+
if self.config.name == 'data':
|
43 |
+
features = datasets.Features({
|
44 |
+
'id': datasets.Value('int32'),
|
45 |
+
'text': datasets.Value('string'),
|
46 |
+
'entities': datasets.Sequence(datasets.Value('string'))
|
47 |
+
})
|
48 |
+
else:
|
49 |
+
features = datasets.Features({
|
50 |
+
'type': datasets.Value('string')
|
51 |
+
})
|
52 |
+
return datasets.DatasetInfo(
|
53 |
+
description=_DESCRIPTION,
|
54 |
+
features=features,
|
55 |
+
homepage=_HOMEPAGE,
|
56 |
+
citation=_CITATION
|
57 |
+
)
|
58 |
+
|
59 |
+
def _split_generators(self, dl_manager: datasets.DownloadManager):
|
60 |
+
if self.config.name == 'data':
|
61 |
+
files = dl_manager.download(self._DATA_URLS)
|
62 |
+
return [
|
63 |
+
datasets.SplitGenerator(
|
64 |
+
name=datasets.Split.TRAIN,
|
65 |
+
gen_kwargs={'filepath': files['train']},
|
66 |
+
),
|
67 |
+
datasets.SplitGenerator(
|
68 |
+
name=datasets.Split.TEST,
|
69 |
+
gen_kwargs={'filepath': files['test']},
|
70 |
+
),
|
71 |
+
datasets.SplitGenerator(
|
72 |
+
name='dev',
|
73 |
+
gen_kwargs={'filepath': files['dev']},
|
74 |
+
),
|
75 |
+
]
|
76 |
+
else:
|
77 |
+
files = dl_manager.download(self._ENTITY_TYPES_URLS)
|
78 |
+
return [datasets.SplitGenerator(
|
79 |
+
name='ent_types',
|
80 |
+
gen_kwargs={'filepath': files['ent_types']},
|
81 |
+
)]
|
82 |
+
|
83 |
+
def _generate_examples(self, filepath):
|
84 |
+
if self.config.name == 'data':
|
85 |
+
with open(filepath, encoding='utf-8') as f:
|
86 |
+
for line in f:
|
87 |
+
doc = json.loads(line)
|
88 |
+
yield doc['id'], doc
|
89 |
+
else:
|
90 |
+
with open(filepath, encoding='utf-8') as f:
|
91 |
+
for i, line in enumerate(f):
|
92 |
+
entity_type = line.strip()
|
93 |
+
if entity_type:
|
94 |
+
yield i, {'type': entity_type}
|
data/dev.jsonl
ADDED
The diff for this file is too large to render.
See raw diff
|
|
data/test.jsonl
ADDED
The diff for this file is too large to render.
See raw diff
|
|
data/train.jsonl
ADDED
The diff for this file is too large to render.
See raw diff
|
|
ent_types.txt
ADDED
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
AGE
|
2 |
+
AWARD
|
3 |
+
CITY
|
4 |
+
COUNTRY
|
5 |
+
CRIME
|
6 |
+
DATE
|
7 |
+
DISEASE
|
8 |
+
DISTRICT
|
9 |
+
EVENT
|
10 |
+
FACILITY
|
11 |
+
FAMILY
|
12 |
+
IDEOLOGY
|
13 |
+
LANGUAGE
|
14 |
+
LAW
|
15 |
+
LOCATION
|
16 |
+
MONEY
|
17 |
+
NATIONALITY
|
18 |
+
NUMBER
|
19 |
+
ORDINAL
|
20 |
+
ORGANIZATION
|
21 |
+
PENALTY
|
22 |
+
PERCENT
|
23 |
+
PERSON
|
24 |
+
PRODUCT
|
25 |
+
PROFESSION
|
26 |
+
RELIGION
|
27 |
+
STATE_OR_PROVINCE
|
28 |
+
TIME
|
29 |
+
WORK_OF_ART
|