holylovenia
commited on
Commit
·
d50c885
1
Parent(s):
db87cea
Upload indqner.py with huggingface_hub
Browse files- indqner.py +181 -0
indqner.py
ADDED
@@ -0,0 +1,181 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
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 |
+
""" IndQNER Dataset """
|
16 |
+
|
17 |
+
from pathlib import Path
|
18 |
+
from typing import List
|
19 |
+
|
20 |
+
import datasets
|
21 |
+
|
22 |
+
from nusacrowd.utils import schemas
|
23 |
+
from nusacrowd.utils.common_parser import load_conll_data
|
24 |
+
from nusacrowd.utils.configs import NusantaraConfig
|
25 |
+
from nusacrowd.utils.constants import Tasks
|
26 |
+
|
27 |
+
_CITATION = """\
|
28 |
+
@misc{,
|
29 |
+
author = {Ria Hari Gusmita, Asep Fajar Firmansyah, Khodijah Khuliyah},
|
30 |
+
title = {{IndQNER: a NER Benchmark Dataset on Indonesian Translation of Quran}},
|
31 |
+
url = {https://github.com/dice-group/IndQNER},
|
32 |
+
year = {2022}
|
33 |
+
}
|
34 |
+
"""
|
35 |
+
|
36 |
+
_LOCAL = False
|
37 |
+
_LANGUAGES = ["ind"]
|
38 |
+
_DATASETNAME = "IndQNER"
|
39 |
+
_DESCRIPTION = """\
|
40 |
+
IndQNER is a NER dataset created by manually annotating the Indonesian translation of Quran text.
|
41 |
+
The dataset contains 18 named entity categories as follow:
|
42 |
+
"Allah": Allah (including synonim of Allah such as Yang maha mengetahui lagi mahabijaksana)
|
43 |
+
"Throne": Throne of Allah (such as 'Arasy)
|
44 |
+
"Artifact": Artifact (such as Ka'bah, Baitullah)
|
45 |
+
"AstronomicalBody": Astronomical body (such as bumi, matahari)
|
46 |
+
"Event": Event (such as hari akhir, kiamat)
|
47 |
+
"HolyBook": Holy book (such as AlQur'an)
|
48 |
+
"Language": Language (such as bahasa Arab
|
49 |
+
"Angel": Angel (such as Jibril, Mikail)
|
50 |
+
"Person": Person (such as Bani Israil, Fir'aun)
|
51 |
+
"Messenger": Messenger (such as Isa, Muhammad, Musa)
|
52 |
+
"Prophet": Prophet (such as Adam, Sulaiman)
|
53 |
+
"AfterlifeLocation": Afterlife location (such as Jahanam, Jahim, Padang Mahsyar)
|
54 |
+
"GeographicalLocation": Geographical location (such as Sinai, negeru Babilonia)
|
55 |
+
"Color": Color (such as kuning tua)
|
56 |
+
"Religion": Religion (such as Islam, Yahudi, Nasrani)
|
57 |
+
"Food": Food (such as manna, salwa)
|
58 |
+
"""
|
59 |
+
|
60 |
+
_HOMEPAGE = "https://github.com/dice-group/IndQNER"
|
61 |
+
_LICENSE = "Unknown"
|
62 |
+
_URLs = {
|
63 |
+
"train": "https://raw.githubusercontent.com/dice-group/IndQNER/master/datasets/train.txt",
|
64 |
+
"validation": "https://raw.githubusercontent.com/dice-group/IndQNER/master/datasets/dev.txt",
|
65 |
+
"test": "https://raw.githubusercontent.com/dice-group/IndQNER/master/datasets/test.txt",
|
66 |
+
}
|
67 |
+
_SUPPORTED_TASKS = [Tasks.NAMED_ENTITY_RECOGNITION]
|
68 |
+
_SOURCE_VERSION = "1.0.0"
|
69 |
+
_NUSANTARA_VERSION = "1.0.0"
|
70 |
+
|
71 |
+
|
72 |
+
class IndqnerDataset(datasets.GeneratorBasedBuilder):
|
73 |
+
"""IndQNER is an Named Entity Recognition benchmark dataset on a niche domain i.e. Indonesian Translation of Quran."""
|
74 |
+
|
75 |
+
label_classes = [
|
76 |
+
"B-Allah",
|
77 |
+
"B-Throne",
|
78 |
+
"B-Artifact",
|
79 |
+
"B-AstronomicalBody",
|
80 |
+
"B-Event",
|
81 |
+
"B-HolyBook",
|
82 |
+
"B-Language",
|
83 |
+
"B-Angel",
|
84 |
+
"B-Person",
|
85 |
+
"B-Messenger",
|
86 |
+
"B-Prophet",
|
87 |
+
"B-AfterlifeLocation",
|
88 |
+
"B-GeographicalLocation",
|
89 |
+
"B-Color",
|
90 |
+
"B-Religion",
|
91 |
+
"B-Food",
|
92 |
+
"I-Allah",
|
93 |
+
"I-Throne",
|
94 |
+
"I-Artifact",
|
95 |
+
"I-AstronomicalBody",
|
96 |
+
"I-Event",
|
97 |
+
"I-HolyBook",
|
98 |
+
"I-Language",
|
99 |
+
"I-Angel",
|
100 |
+
"I-Person",
|
101 |
+
"I-Messenger",
|
102 |
+
"I-Prophet",
|
103 |
+
"I-AfterlifeLocation",
|
104 |
+
"I-GeographicalLocation",
|
105 |
+
"I-Color",
|
106 |
+
"I-Religion",
|
107 |
+
"I-Food",
|
108 |
+
"O",
|
109 |
+
]
|
110 |
+
|
111 |
+
BUILDER_CONFIGS = [
|
112 |
+
NusantaraConfig(
|
113 |
+
name="indqner_source",
|
114 |
+
version=datasets.Version(_SOURCE_VERSION),
|
115 |
+
description="NER dataset from Indonesian translation Quran source schema",
|
116 |
+
schema="source",
|
117 |
+
subset_id="indqner",
|
118 |
+
),
|
119 |
+
NusantaraConfig(
|
120 |
+
name="indqner_nusantara_seq_label",
|
121 |
+
version=datasets.Version(_SOURCE_VERSION),
|
122 |
+
description="NER dataset from Indonesian translation Quran Nusantara schema",
|
123 |
+
schema="nusantara_seq_label",
|
124 |
+
subset_id="indqner",
|
125 |
+
),
|
126 |
+
]
|
127 |
+
|
128 |
+
DEFAULT_CONFIG_NAME = "indqner_source"
|
129 |
+
|
130 |
+
def _info(self):
|
131 |
+
if self.config.schema == "source":
|
132 |
+
features = datasets.Features({"index": datasets.Value("string"), "tokens": [datasets.Value("string")], "ner_tag": [datasets.Value("string")]})
|
133 |
+
elif self.config.schema == "nusantara_seq_label":
|
134 |
+
features = schemas.seq_label_features(self.label_classes)
|
135 |
+
|
136 |
+
return datasets.DatasetInfo(
|
137 |
+
description=_DESCRIPTION,
|
138 |
+
features=features,
|
139 |
+
homepage=_HOMEPAGE,
|
140 |
+
license=_LICENSE,
|
141 |
+
citation=_CITATION,
|
142 |
+
)
|
143 |
+
|
144 |
+
def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]:
|
145 |
+
train_tsv_path = Path(dl_manager.download_and_extract(_URLs["train"]))
|
146 |
+
validation_tsv_path = Path(dl_manager.download_and_extract(_URLs["validation"]))
|
147 |
+
test_tsv_path = Path(dl_manager.download_and_extract(_URLs["test"]))
|
148 |
+
data_files = {
|
149 |
+
"train": train_tsv_path,
|
150 |
+
"validation": validation_tsv_path,
|
151 |
+
"test": test_tsv_path,
|
152 |
+
}
|
153 |
+
|
154 |
+
return [
|
155 |
+
datasets.SplitGenerator(
|
156 |
+
name=datasets.Split.TRAIN,
|
157 |
+
gen_kwargs={"filepath": data_files["train"]},
|
158 |
+
),
|
159 |
+
datasets.SplitGenerator(
|
160 |
+
name=datasets.Split.VALIDATION,
|
161 |
+
gen_kwargs={"filepath": data_files["validation"]},
|
162 |
+
),
|
163 |
+
datasets.SplitGenerator(
|
164 |
+
name=datasets.Split.TEST,
|
165 |
+
gen_kwargs={"filepath": data_files["test"]},
|
166 |
+
),
|
167 |
+
]
|
168 |
+
|
169 |
+
def _generate_examples(self, filepath: Path):
|
170 |
+
conll_dataset = load_conll_data(filepath)
|
171 |
+
|
172 |
+
if self.config.schema == "source":
|
173 |
+
for index, row in enumerate(conll_dataset):
|
174 |
+
ex = {"index": str(index), "tokens": row["sentence"], "ner_tag": row["label"]}
|
175 |
+
yield index, ex
|
176 |
+
elif self.config.schema == "nusantara_seq_label":
|
177 |
+
for index, row in enumerate(conll_dataset):
|
178 |
+
ex = {"id": str(index), "tokens": row["sentence"], "labels": row["label"]}
|
179 |
+
yield index, ex
|
180 |
+
else:
|
181 |
+
raise ValueError(f"Invalid config: {self.config.name}")
|