Commit
·
468ebd1
1
Parent(s):
b517279
upload hubscripts/meddocan_hub.py to hub from bigbio repo
Browse files- meddocan.py +249 -0
meddocan.py
ADDED
@@ -0,0 +1,249 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# coding=utf-8
|
2 |
+
# Copyright 2022 The HuggingFace Datasets Authors and the current dataset script contributor.
|
3 |
+
#
|
4 |
+
# Licensed under the Apache License, Version 2.0 (the "License");
|
5 |
+
# you may not use this file except in compliance with the License.
|
6 |
+
# You may obtain a copy of the License at
|
7 |
+
#
|
8 |
+
# http://www.apache.org/licenses/LICENSE-2.0
|
9 |
+
#
|
10 |
+
# Unless required by applicable law or agreed to in writing, software
|
11 |
+
# distributed under the License is distributed on an "AS IS" BASIS,
|
12 |
+
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
13 |
+
# See the License for the specific language governing permissions and
|
14 |
+
# limitations under the License.
|
15 |
+
|
16 |
+
"""
|
17 |
+
A dataset loading script for the MEDDOCAN corpus.
|
18 |
+
The MEDDOCAN datset is a manually annotated collection of clinical case
|
19 |
+
reports derived from the Spanish Clinical Case Corpus (SPACCC). It was designed
|
20 |
+
for the Medical Document Anonymization Track, the first the first community
|
21 |
+
challenge task specifically devoted to the anonymization of medical documents in Spanish
|
22 |
+
"""
|
23 |
+
|
24 |
+
import os
|
25 |
+
from pathlib import Path
|
26 |
+
from typing import Dict, List, Tuple
|
27 |
+
|
28 |
+
import datasets
|
29 |
+
|
30 |
+
from .bigbiohub import kb_features
|
31 |
+
from .bigbiohub import BigBioConfig
|
32 |
+
from .bigbiohub import Tasks
|
33 |
+
|
34 |
+
_LANGUAGES = ['Spanish']
|
35 |
+
_PUBMED = False
|
36 |
+
_LOCAL = False
|
37 |
+
_CITATION = """\
|
38 |
+
@inproceedings{marimon2019automatic,
|
39 |
+
title={Automatic De-identification of Medical Texts in Spanish: the MEDDOCAN Track, Corpus, Guidelines, Methods and Evaluation of Results.},
|
40 |
+
author={Marimon, Montserrat and Gonzalez-Agirre, Aitor and Intxaurrondo, Ander and Rodriguez, Heidy and Martin, Jose Lopez and Villegas, Marta and Krallinger, Martin},
|
41 |
+
booktitle={IberLEF@ SEPLN},
|
42 |
+
pages={618--638},
|
43 |
+
year={2019}
|
44 |
+
}
|
45 |
+
"""
|
46 |
+
|
47 |
+
_DATASETNAME = "meddocan"
|
48 |
+
_DISPLAYNAME = "MEDDOCAN"
|
49 |
+
|
50 |
+
_DESCRIPTION = """\
|
51 |
+
MEDDOCAN: Medical Document Anonymization Track
|
52 |
+
|
53 |
+
This dataset is designed for the MEDDOCAN task, sponsored by Plan de Impulso de las Tecnologías del Lenguaje.
|
54 |
+
|
55 |
+
It is a manually classified collection of 1,000 clinical case reports derived from the \
|
56 |
+
Spanish Clinical Case Corpus (SPACCC), enriched with PHI expressions.
|
57 |
+
|
58 |
+
The annotation of the entire set of entity mentions was carried out by experts annotators\
|
59 |
+
and it includes 29 entity types relevant for the annonymiation of medical documents.\
|
60 |
+
22 of these annotation types are actually present in the corpus: TERRITORIO, FECHAS, \
|
61 |
+
EDAD_SUJETO_ASISTENCIA, NOMBRE_SUJETO_ASISTENCIA, NOMBRE_PERSONAL_SANITARIO, \
|
62 |
+
SEXO_SUJETO_ASISTENCIA, CALLE, PAIS, ID_SUJETO_ASISTENCIA, CORREO, ID_TITULACION_PERSONAL_SANITARIO,\
|
63 |
+
ID_ASEGURAMIENTO, HOSPITAL, FAMILIARES_SUJETO_ASISTENCIA, INSTITUCION, ID_CONTACTO ASISTENCIAL,\
|
64 |
+
NUMERO_TELEFONO, PROFESION, NUMERO_FAX, OTROS_SUJETO_ASISTENCIA, CENTRO_SALUD, ID_EMPLEO_PERSONAL_SANITARIO
|
65 |
+
|
66 |
+
For further information, please visit https://temu.bsc.es/meddocan/ or send an email to [email protected]
|
67 |
+
"""
|
68 |
+
|
69 |
+
|
70 |
+
_HOMEPAGE = "https://temu.bsc.es/meddocan/"
|
71 |
+
|
72 |
+
_LICENSE = 'Creative Commons Attribution 4.0 International'
|
73 |
+
|
74 |
+
_URLS = {
|
75 |
+
"meddocan": "https://zenodo.org/record/4279323/files/meddocan.zip?download=1",
|
76 |
+
}
|
77 |
+
|
78 |
+
_SUPPORTED_TASKS = [Tasks.NAMED_ENTITY_RECOGNITION]
|
79 |
+
|
80 |
+
_SOURCE_VERSION = "1.0.0"
|
81 |
+
|
82 |
+
_BIGBIO_VERSION = "1.0.0"
|
83 |
+
|
84 |
+
|
85 |
+
class MeddocanDataset(datasets.GeneratorBasedBuilder):
|
86 |
+
"""Manually annotated collection of clinical case studies from Spanish medical publications."""
|
87 |
+
|
88 |
+
SOURCE_VERSION = datasets.Version(_SOURCE_VERSION)
|
89 |
+
BIGBIO_VERSION = datasets.Version(_BIGBIO_VERSION)
|
90 |
+
|
91 |
+
BUILDER_CONFIGS = [
|
92 |
+
BigBioConfig(
|
93 |
+
name="meddocan_source",
|
94 |
+
version=SOURCE_VERSION,
|
95 |
+
description="Meddocan source schema",
|
96 |
+
schema="source",
|
97 |
+
subset_id="meddocan",
|
98 |
+
),
|
99 |
+
BigBioConfig(
|
100 |
+
name="meddocan_bigbio_kb",
|
101 |
+
version=BIGBIO_VERSION,
|
102 |
+
description="Meddocan BigBio schema",
|
103 |
+
schema="bigbio_kb",
|
104 |
+
subset_id="meddocan",
|
105 |
+
),
|
106 |
+
]
|
107 |
+
|
108 |
+
DEFAULT_CONFIG_NAME = "meddocan_source"
|
109 |
+
|
110 |
+
def _info(self) -> datasets.DatasetInfo:
|
111 |
+
if self.config.schema == "source":
|
112 |
+
features = datasets.Features(
|
113 |
+
{
|
114 |
+
"id": datasets.Value("string"),
|
115 |
+
"document_id": datasets.Value("string"),
|
116 |
+
"text": datasets.Value("string"),
|
117 |
+
# "labels": [datasets.Value("string")],
|
118 |
+
"text_bound_annotations": [ # T line in brat
|
119 |
+
{
|
120 |
+
"offsets": datasets.Sequence([datasets.Value("int32")]),
|
121 |
+
"text": datasets.Sequence(datasets.Value("string")),
|
122 |
+
"type": datasets.Value("string"),
|
123 |
+
"id": datasets.Value("string"),
|
124 |
+
}
|
125 |
+
],
|
126 |
+
"events": [ # E line in brat
|
127 |
+
{
|
128 |
+
"trigger": datasets.Value("string"),
|
129 |
+
"id": datasets.Value("string"),
|
130 |
+
"type": datasets.Value("string"),
|
131 |
+
"arguments": datasets.Sequence(
|
132 |
+
{
|
133 |
+
"role": datasets.Value("string"),
|
134 |
+
"ref_id": datasets.Value("string"),
|
135 |
+
}
|
136 |
+
),
|
137 |
+
}
|
138 |
+
],
|
139 |
+
"relations": [ # R line in brat
|
140 |
+
{
|
141 |
+
"id": datasets.Value("string"),
|
142 |
+
"head": {
|
143 |
+
"ref_id": datasets.Value("string"),
|
144 |
+
"role": datasets.Value("string"),
|
145 |
+
},
|
146 |
+
"tail": {
|
147 |
+
"ref_id": datasets.Value("string"),
|
148 |
+
"role": datasets.Value("string"),
|
149 |
+
},
|
150 |
+
"type": datasets.Value("string"),
|
151 |
+
}
|
152 |
+
],
|
153 |
+
"equivalences": [ # Equiv line in brat
|
154 |
+
{
|
155 |
+
"id": datasets.Value("string"),
|
156 |
+
"ref_ids": datasets.Sequence(datasets.Value("string")),
|
157 |
+
}
|
158 |
+
],
|
159 |
+
"attributes": [ # M or A lines in brat
|
160 |
+
{
|
161 |
+
"id": datasets.Value("string"),
|
162 |
+
"type": datasets.Value("string"),
|
163 |
+
"ref_id": datasets.Value("string"),
|
164 |
+
"value": datasets.Value("string"),
|
165 |
+
}
|
166 |
+
],
|
167 |
+
"normalizations": [ # N lines in brat
|
168 |
+
{
|
169 |
+
"id": datasets.Value("string"),
|
170 |
+
"type": datasets.Value("string"),
|
171 |
+
"ref_id": datasets.Value("string"),
|
172 |
+
"resource_name": datasets.Value("string"),
|
173 |
+
"cuid": datasets.Value("string"),
|
174 |
+
"text": datasets.Value("string"),
|
175 |
+
}
|
176 |
+
],
|
177 |
+
},
|
178 |
+
)
|
179 |
+
|
180 |
+
elif self.config.schema == "bigbio_kb":
|
181 |
+
features = kb_features
|
182 |
+
|
183 |
+
return datasets.DatasetInfo(
|
184 |
+
description=_DESCRIPTION,
|
185 |
+
features=features,
|
186 |
+
homepage=_HOMEPAGE,
|
187 |
+
license=str(_LICENSE),
|
188 |
+
citation=_CITATION,
|
189 |
+
)
|
190 |
+
|
191 |
+
def _split_generators(self, dl_manager) -> List[datasets.SplitGenerator]:
|
192 |
+
"""
|
193 |
+
Downloads/extracts the data to generate the train, validation and test splits.
|
194 |
+
Each split is created by instantiating a `datasets.SplitGenerator`, which will
|
195 |
+
call `this._generate_examples` with the keyword arguments in `gen_kwargs`.
|
196 |
+
"""
|
197 |
+
|
198 |
+
data_dir = dl_manager.download_and_extract(_URLS["meddocan"])
|
199 |
+
|
200 |
+
return [
|
201 |
+
datasets.SplitGenerator(
|
202 |
+
name=datasets.Split.TRAIN,
|
203 |
+
gen_kwargs={
|
204 |
+
"filepath": Path(os.path.join(data_dir, "meddocan/train/brat")),
|
205 |
+
"split": "train",
|
206 |
+
},
|
207 |
+
),
|
208 |
+
datasets.SplitGenerator(
|
209 |
+
name=datasets.Split.TEST,
|
210 |
+
gen_kwargs={
|
211 |
+
"filepath": Path(os.path.join(data_dir, "meddocan/test/brat")),
|
212 |
+
"split": "test",
|
213 |
+
},
|
214 |
+
),
|
215 |
+
datasets.SplitGenerator(
|
216 |
+
name=datasets.Split.VALIDATION,
|
217 |
+
gen_kwargs={
|
218 |
+
"filepath": Path(os.path.join(data_dir, "meddocan/dev/brat")),
|
219 |
+
"split": "dev",
|
220 |
+
},
|
221 |
+
),
|
222 |
+
]
|
223 |
+
|
224 |
+
def _generate_examples(self, filepath, split: str) -> Tuple[int, Dict]:
|
225 |
+
"""
|
226 |
+
This method handles input defined in _split_generators to yield (key, example) tuples from the dataset.
|
227 |
+
Method parameters are unpacked from `gen_kwargs` as given in `_split_generators`.
|
228 |
+
"""
|
229 |
+
|
230 |
+
txt_files = sorted(list(filepath.glob("*txt")))
|
231 |
+
# tsv_files = sorted(list(filepaths[1].glob("*tsv")))
|
232 |
+
|
233 |
+
if self.config.schema == "source":
|
234 |
+
for guid, txt_file in enumerate(txt_files):
|
235 |
+
example = parsing.parse_brat_file(txt_file)
|
236 |
+
|
237 |
+
example["id"] = str(guid)
|
238 |
+
yield guid, example
|
239 |
+
|
240 |
+
elif self.config.schema == "bigbio_kb":
|
241 |
+
for guid, txt_file in enumerate(txt_files):
|
242 |
+
example = parsing.brat_parse_to_bigbio_kb(
|
243 |
+
parsing.parse_brat_file(txt_file)
|
244 |
+
)
|
245 |
+
example["id"] = str(guid)
|
246 |
+
yield guid, example
|
247 |
+
|
248 |
+
else:
|
249 |
+
raise ValueError(f"Invalid config: {self.config.name}")
|