Commit
·
d5d71fd
1
Parent(s):
c63bdb3
upload hubscripts/paramed_hub.py to hub from bigbio repo
Browse files- paramed.py +197 -0
paramed.py
ADDED
@@ -0,0 +1,197 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# coding=utf-8
|
2 |
+
# Copyright 2020 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 |
+
NEJM is a Chinese-English parallel corpus crawled from the New England Journal of Medicine website.
|
18 |
+
English articles are distributed through https://www.nejm.org/ and Chinese articles are distributed through
|
19 |
+
http://nejmqianyan.cn/. The corpus contains all article pairs (around 2000 pairs) since 2011.
|
20 |
+
The script loads dataset in bigbio schema (using schemas/text-to-text) AND/OR source (default) schema
|
21 |
+
"""
|
22 |
+
import os # useful for paths
|
23 |
+
from typing import Dict, Iterable, List
|
24 |
+
|
25 |
+
import datasets
|
26 |
+
|
27 |
+
from .bigbiohub import text2text_features
|
28 |
+
from .bigbiohub import BigBioConfig
|
29 |
+
from .bigbiohub import Tasks
|
30 |
+
|
31 |
+
logger = datasets.logging.get_logger(__name__)
|
32 |
+
|
33 |
+
|
34 |
+
_LANGUAGES = ['English', 'Chinese']
|
35 |
+
_PUBMED = False
|
36 |
+
_LOCAL = False
|
37 |
+
_CITATION = """\
|
38 |
+
@article{liu2021paramed,
|
39 |
+
author = {Liu, Boxiang and Huang, Liang},
|
40 |
+
title = {ParaMed: a parallel corpus for English–Chinese translation in the biomedical domain},
|
41 |
+
journal = {BMC Medical Informatics and Decision Making},
|
42 |
+
volume = {21},
|
43 |
+
year = {2021},
|
44 |
+
url = {https://bmcmedinformdecismak.biomedcentral.com/articles/10.1186/s12911-021-01621-8},
|
45 |
+
doi = {10.1186/s12911-021-01621-8}
|
46 |
+
}
|
47 |
+
"""
|
48 |
+
_DATASETNAME = "paramed"
|
49 |
+
_DISPLAYNAME = "ParaMed"
|
50 |
+
|
51 |
+
_DESCRIPTION = """\
|
52 |
+
NEJM is a Chinese-English parallel corpus crawled from the New England Journal of Medicine website.
|
53 |
+
English articles are distributed through https://www.nejm.org/ and Chinese articles are distributed through
|
54 |
+
http://nejmqianyan.cn/. The corpus contains all article pairs (around 2000 pairs) since 2011.
|
55 |
+
"""
|
56 |
+
|
57 |
+
_HOMEPAGE = "https://github.com/boxiangliu/ParaMed"
|
58 |
+
|
59 |
+
_LICENSE = 'Creative Commons Attribution 4.0 International'
|
60 |
+
|
61 |
+
_URLs = {
|
62 |
+
"source": "https://github.com/boxiangliu/ParaMed/blob/master/data/nejm-open-access.tar.gz?raw=true",
|
63 |
+
"bigbio_t2t": "https://github.com/boxiangliu/ParaMed/blob/master/data/nejm-open-access.tar.gz?raw=true",
|
64 |
+
}
|
65 |
+
_SUPPORTED_TASKS = [Tasks.TRANSLATION]
|
66 |
+
_SOURCE_VERSION = "1.0.0"
|
67 |
+
_BIGBIO_VERSION = "1.0.0"
|
68 |
+
|
69 |
+
_DATA_DIR = "./processed_data/open_access/open_access"
|
70 |
+
|
71 |
+
|
72 |
+
class ParamedDataset(datasets.GeneratorBasedBuilder):
|
73 |
+
"""Write a short docstring documenting what this dataset is"""
|
74 |
+
|
75 |
+
SOURCE_VERSION = datasets.Version(_SOURCE_VERSION)
|
76 |
+
BIGBIO_VERSION = datasets.Version(_BIGBIO_VERSION)
|
77 |
+
|
78 |
+
BUILDER_CONFIGS = [
|
79 |
+
BigBioConfig(
|
80 |
+
name="paramed_source",
|
81 |
+
version=SOURCE_VERSION,
|
82 |
+
description="Paramed source schema",
|
83 |
+
schema="source",
|
84 |
+
subset_id="paramed",
|
85 |
+
),
|
86 |
+
BigBioConfig(
|
87 |
+
name="paramed_bigbio_t2t",
|
88 |
+
version=BIGBIO_VERSION,
|
89 |
+
description="Paramed BigBio schema",
|
90 |
+
schema="bigbio_t2t",
|
91 |
+
subset_id="paramed",
|
92 |
+
),
|
93 |
+
]
|
94 |
+
|
95 |
+
DEFAULT_CONFIG_NAME = "paramed_source"
|
96 |
+
|
97 |
+
def _info(self):
|
98 |
+
|
99 |
+
if self.config.schema == "source":
|
100 |
+
features = datasets.Features(
|
101 |
+
{
|
102 |
+
"document_id": datasets.Value("string"),
|
103 |
+
"text_1": datasets.Value("string"),
|
104 |
+
"text_2": datasets.Value("string"),
|
105 |
+
"text_1_name": datasets.Value("string"),
|
106 |
+
"text_2_name": datasets.Value("string"),
|
107 |
+
}
|
108 |
+
)
|
109 |
+
|
110 |
+
elif self.config.schema == "bigbio_t2t":
|
111 |
+
features = text2text_features
|
112 |
+
|
113 |
+
return datasets.DatasetInfo(
|
114 |
+
# This is the description that will appear on the datasets page.
|
115 |
+
description=_DESCRIPTION,
|
116 |
+
features=features,
|
117 |
+
homepage=_HOMEPAGE,
|
118 |
+
license=str(_LICENSE),
|
119 |
+
citation=_CITATION,
|
120 |
+
)
|
121 |
+
|
122 |
+
def _split_generators(
|
123 |
+
self, dl_manager: datasets.DownloadManager
|
124 |
+
) -> List[datasets.SplitGenerator]:
|
125 |
+
|
126 |
+
my_urls = _URLs[self.config.schema]
|
127 |
+
data_dir = os.path.join(dl_manager.download_and_extract(my_urls), _DATA_DIR)
|
128 |
+
print(data_dir)
|
129 |
+
|
130 |
+
return [
|
131 |
+
datasets.SplitGenerator(
|
132 |
+
name=datasets.Split.TRAIN,
|
133 |
+
gen_kwargs={
|
134 |
+
"filepath": data_dir,
|
135 |
+
"zh_file": os.path.join(data_dir, "nejm.train.zh"),
|
136 |
+
"en_file": os.path.join(data_dir, "nejm.train.en"),
|
137 |
+
"split": "train",
|
138 |
+
},
|
139 |
+
),
|
140 |
+
datasets.SplitGenerator(
|
141 |
+
name=datasets.Split.VALIDATION,
|
142 |
+
gen_kwargs={
|
143 |
+
"filepath": data_dir,
|
144 |
+
"zh_file": os.path.join(data_dir, "nejm.dev.zh"),
|
145 |
+
"en_file": os.path.join(data_dir, "nejm.dev.en"),
|
146 |
+
"split": "dev",
|
147 |
+
},
|
148 |
+
),
|
149 |
+
datasets.SplitGenerator(
|
150 |
+
name=datasets.Split.TEST,
|
151 |
+
gen_kwargs={
|
152 |
+
"filepath": data_dir,
|
153 |
+
"zh_file": os.path.join(data_dir, "nejm.test.zh"),
|
154 |
+
"en_file": os.path.join(data_dir, "nejm.test.en"),
|
155 |
+
"split": "test",
|
156 |
+
},
|
157 |
+
),
|
158 |
+
]
|
159 |
+
|
160 |
+
def _generate_examples(self, filepath, zh_file, en_file, split):
|
161 |
+
|
162 |
+
logger.info("generating examples from = %s", filepath)
|
163 |
+
zh_file = open(zh_file, "r")
|
164 |
+
en_file = open(en_file, "r")
|
165 |
+
zh_file.seek(0)
|
166 |
+
en_file.seek(0)
|
167 |
+
zh_lines = zh_file.readlines()
|
168 |
+
en_lines = en_file.readlines()
|
169 |
+
|
170 |
+
assert len(en_lines) == len(zh_lines), "Line mismatch"
|
171 |
+
|
172 |
+
if self.config.schema == "source":
|
173 |
+
for key, (zh_line, en_line) in enumerate(zip(zh_lines, en_lines)):
|
174 |
+
yield key, {
|
175 |
+
"document_id": str(key),
|
176 |
+
"text_1": zh_line,
|
177 |
+
"text_2": en_line,
|
178 |
+
"text_1_name": "zh",
|
179 |
+
"text_2_name": "en",
|
180 |
+
}
|
181 |
+
zh_file.close()
|
182 |
+
en_file.close()
|
183 |
+
|
184 |
+
elif self.config.schema == "bigbio_t2t":
|
185 |
+
uid = 0
|
186 |
+
for key, (zh_line, en_line) in enumerate(zip(zh_lines, en_lines)):
|
187 |
+
uid += 1
|
188 |
+
yield key, {
|
189 |
+
"id": str(uid),
|
190 |
+
"document_id": str(key),
|
191 |
+
"text_1": zh_line,
|
192 |
+
"text_2": en_line,
|
193 |
+
"text_1_name": "zh",
|
194 |
+
"text_2_name": "en",
|
195 |
+
}
|
196 |
+
zh_file.close()
|
197 |
+
en_file.close()
|