holylovenia commited on
Commit
f26b48d
1 Parent(s): 3875744

Upload tmad_malay_corpus.py with huggingface_hub

Browse files
Files changed (1) hide show
  1. tmad_malay_corpus.py +140 -0
tmad_malay_corpus.py ADDED
@@ -0,0 +1,140 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ The Towards Malay Abbreviation Disambiguation (TMAD) Malay Corpus includes sentences from Malay news sites with abbreviations and their meanings. Only abbreviations with more than one possible meaning are included.
18
+ """
19
+ import csv
20
+ import json
21
+ from pathlib import Path
22
+ from typing import Dict, List, Tuple
23
+
24
+ import datasets
25
+
26
+ from seacrowd.utils import schemas
27
+ from seacrowd.utils.configs import SEACrowdConfig
28
+ from seacrowd.utils.constants import Licenses, Tasks
29
+
30
+ _CITATION = """\
31
+ @article{article,
32
+ author = {Ciosici, Manuel and Sommer, Tobias},
33
+ year = {2019},
34
+ month = {04},
35
+ pages = {},
36
+ title = {Unsupervised Abbreviation Disambiguation Contextual disambiguation using word embeddings}
37
+ }
38
+ """
39
+
40
+ _DATASETNAME = "tmad_malay_corpus"
41
+
42
+ _DESCRIPTION = """\
43
+ The Towards Malay Abbreviation Disambiguation (TMAD) Malay Corpus includes sentences from Malay news sites with abbreviations and their meanings. Only abbreviations with more than one possible meaning are included.
44
+ """
45
+
46
+ _HOMEPAGE = "https://github.com/bhysss/TMAD-CUM/tree/master"
47
+
48
+ _LANGUAGES = ["zlm"]
49
+
50
+ _LICENSE = Licenses.UNKNOWN.value
51
+
52
+ _LOCAL = False
53
+
54
+ _URLS = {
55
+ "train": "https://raw.githubusercontent.com/bhysss/TMAD-CUM/master/data/Malay/data_train.csv",
56
+ "dev": "https://raw.githubusercontent.com/bhysss/TMAD-CUM/master/data/Malay/data_dev.csv",
57
+ "test": "https://raw.githubusercontent.com/bhysss/TMAD-CUM/master/data/Malay/data_test.csv",
58
+ "dict": "https://raw.githubusercontent.com/bhysss/TMAD-CUM/master/data/Malay/May_dic.json",
59
+ }
60
+ _SUPPORTED_TASKS = [Tasks.QUESTION_ANSWERING]
61
+
62
+ _SOURCE_VERSION = "1.0.0"
63
+ _SEACROWD_VERSION = "2024.06.20"
64
+
65
+
66
+ class TMADMalayCorpusDataset(datasets.GeneratorBasedBuilder):
67
+ """Abbreviation disambiguation dataset from Malay news sites."""
68
+
69
+ SOURCE_VERSION = datasets.Version(_SOURCE_VERSION)
70
+ SEACROWD_VERSION = datasets.Version(_SEACROWD_VERSION)
71
+
72
+ BUILDER_CONFIGS = [
73
+ SEACrowdConfig(
74
+ name=f"{_DATASETNAME}_source",
75
+ version=SOURCE_VERSION,
76
+ description="{_DATASETNAME} source schema",
77
+ schema="source",
78
+ subset_id=f"{_DATASETNAME}",
79
+ ),
80
+ SEACrowdConfig(
81
+ name=f"{_DATASETNAME}_seacrowd_qa",
82
+ version=SEACROWD_VERSION,
83
+ description=f"{_DATASETNAME} SEACrowd schema",
84
+ schema="seacrowd_qa",
85
+ subset_id=f"{_DATASETNAME}",
86
+ ),
87
+ ]
88
+
89
+ DEFAULT_CONFIG_NAME = f"{_DATASETNAME}_source"
90
+
91
+ def _info(self) -> datasets.DatasetInfo:
92
+ if self.config.schema == "source":
93
+ features = datasets.Features({"abbr": datasets.Value("string"), "definition": datasets.Value("string"), "sentence": datasets.Value("string"), "choices": datasets.Sequence(datasets.Value("string"))})
94
+
95
+ elif self.config.schema == "seacrowd_qa":
96
+ features = schemas.qa_features
97
+
98
+ return datasets.DatasetInfo(
99
+ description=_DESCRIPTION,
100
+ features=features,
101
+ homepage=_HOMEPAGE,
102
+ license=_LICENSE,
103
+ citation=_CITATION,
104
+ )
105
+
106
+ def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]:
107
+ """Returns SplitGenerators."""
108
+
109
+ data_dirs = dl_manager.download_and_extract(_URLS)
110
+
111
+ return [
112
+ datasets.SplitGenerator(
113
+ name=datasets.Split.TRAIN,
114
+ # Whatever you put in gen_kwargs will be passed to _generate_examples
115
+ gen_kwargs={"filepath": data_dirs["train"], "dictpath": data_dirs["dict"]},
116
+ ),
117
+ datasets.SplitGenerator(
118
+ name=datasets.Split.TEST,
119
+ gen_kwargs={"filepath": data_dirs["test"], "dictpath": data_dirs["dict"]},
120
+ ),
121
+ datasets.SplitGenerator(
122
+ name=datasets.Split.VALIDATION,
123
+ gen_kwargs={"filepath": data_dirs["dev"], "dictpath": data_dirs["dict"]},
124
+ ),
125
+ ]
126
+
127
+ def _generate_examples(self, filepath: Path, dictpath: Path) -> Tuple[int, Dict]:
128
+
129
+ with open(dictpath) as f:
130
+ may_dict = json.load(f)
131
+
132
+ if self.config.schema == "source":
133
+ with open(filepath, encoding="utf-8") as f:
134
+ for row_idx, row in enumerate(csv.DictReader(f)):
135
+ yield row_idx, {"abbr": row["Abbr"], "definition": row["Definition"], "sentence": row["Sentence"], "choices": may_dict[row["Abbr"]]}
136
+
137
+ elif self.config.schema == "seacrowd_qa":
138
+ with open(filepath, encoding="utf-8") as f:
139
+ for row_idx, row in enumerate(csv.DictReader(f)):
140
+ yield row_idx, {"id": row_idx, "question_id": 0, "document_id": 0, "question": row["Abbr"], "type": "multiple_choice", "choices": may_dict[row["Abbr"]], "context": row["Sentence"], "answer": [row["Definition"]], "meta": {}}