holylovenia
commited on
Commit
•
06df0f5
1
Parent(s):
a09a6f7
Upload indocollex.py with huggingface_hub
Browse files- indocollex.py +189 -0
indocollex.py
ADDED
@@ -0,0 +1,189 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
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 |
+
from itertools import chain
|
17 |
+
from pathlib import Path
|
18 |
+
from typing import Dict, List, Tuple
|
19 |
+
|
20 |
+
import datasets
|
21 |
+
|
22 |
+
from nusacrowd.utils import schemas
|
23 |
+
from nusacrowd.utils.configs import NusantaraConfig
|
24 |
+
from nusacrowd.utils.constants import Tasks
|
25 |
+
|
26 |
+
_CITATION = """\
|
27 |
+
@inproceedings{wibowo-etal-2021-indocollex,
|
28 |
+
title = "{I}ndo{C}ollex: A Testbed for Morphological Transformation of {I}ndonesian Word Colloquialism",
|
29 |
+
author = {Wibowo, Haryo Akbarianto and Nityasya, Made Nindyatama and Aky{\"u}rek, Afra Feyza and Fitriany, Suci and Aji, Alham Fikri and Prasojo, Radityo Eko and Wijaya, Derry Tanti},
|
30 |
+
booktitle = "Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021",
|
31 |
+
month = aug,
|
32 |
+
year = "2021",
|
33 |
+
address = "Online",
|
34 |
+
publisher = "Association for Computational Linguistics",
|
35 |
+
url = "https://aclanthology.org/2021.findings-acl.280",
|
36 |
+
doi = "10.18653/v1/2021.findings-acl.280",
|
37 |
+
pages = "3170--3183",
|
38 |
+
}"""
|
39 |
+
|
40 |
+
_LANGUAGES = ["ind"]
|
41 |
+
_LOCAL = False
|
42 |
+
|
43 |
+
_DATASETNAME = "indocollex"
|
44 |
+
|
45 |
+
_DESCRIPTION = """\
|
46 |
+
IndoCollex: A Testbed for Morphological Transformation of Indonesian Colloquial Words
|
47 |
+
"""
|
48 |
+
|
49 |
+
_HOMEPAGE = "https://github.com/haryoa/indo-collex"
|
50 |
+
|
51 |
+
_LICENSE = "CC BY-SA 4.0"
|
52 |
+
|
53 |
+
_URLS = {
|
54 |
+
_DATASETNAME: {
|
55 |
+
"train": "https://github.com/haryoa/indo-collex/raw/main/data/full.csv",
|
56 |
+
},
|
57 |
+
f"{_DATASETNAME}_f2i": {
|
58 |
+
"train": "https://github.com/haryoa/indo-collex/raw/main/data/formal_to_informal/train.csv",
|
59 |
+
"dev": "https://github.com/haryoa/indo-collex/raw/main/data/formal_to_informal/dev.csv",
|
60 |
+
"test": "https://github.com/haryoa/indo-collex/raw/main/data/formal_to_informal/test.csv",
|
61 |
+
},
|
62 |
+
f"{_DATASETNAME}_i2f": {
|
63 |
+
"train": "https://github.com/haryoa/indo-collex/raw/main/data/informal_to_formal/train.csv",
|
64 |
+
"dev": "https://github.com/haryoa/indo-collex/raw/main/data/informal_to_formal/dev.csv",
|
65 |
+
"test": "https://github.com/haryoa/indo-collex/raw/main/data/informal_to_formal/test.csv",
|
66 |
+
},
|
67 |
+
}
|
68 |
+
|
69 |
+
_SUPPORTED_TASKS = [Tasks.MORPHOLOGICAL_INFLECTION]
|
70 |
+
|
71 |
+
_SOURCE_VERSION = "1.0.0"
|
72 |
+
_NUSANTARA_VERSION = "1.0.0"
|
73 |
+
|
74 |
+
|
75 |
+
class NewDataset(datasets.GeneratorBasedBuilder):
|
76 |
+
"""IndoCollex: A Testbed for Morphological Transformation of Indonesian Colloquial Words"""
|
77 |
+
|
78 |
+
label_classes = ["acronym", "affixation", "disemvoweling", "rev", "shorten", "sound-alter", "space-dash"]
|
79 |
+
|
80 |
+
BUILDER_CONFIGS = list(
|
81 |
+
chain(
|
82 |
+
*[
|
83 |
+
[
|
84 |
+
NusantaraConfig(
|
85 |
+
name=f"{_DATASETNAME}{suffix}_source",
|
86 |
+
version=datasets.Version(_SOURCE_VERSION),
|
87 |
+
description=f"{_DATASETNAME} source schema",
|
88 |
+
schema="source",
|
89 |
+
subset_id=f"{_DATASETNAME}{suffix}",
|
90 |
+
),
|
91 |
+
NusantaraConfig(
|
92 |
+
name=f"{_DATASETNAME}{suffix}_nusantara_pairs_multi",
|
93 |
+
version=datasets.Version(_NUSANTARA_VERSION),
|
94 |
+
description=f"{_DATASETNAME} Nusantara schema",
|
95 |
+
schema="nusantara_pairs_multi",
|
96 |
+
subset_id=f"{_DATASETNAME}{suffix}",
|
97 |
+
),
|
98 |
+
]
|
99 |
+
for suffix in ["", "_f2i", "_i2f"]
|
100 |
+
]
|
101 |
+
)
|
102 |
+
)
|
103 |
+
|
104 |
+
DEFAULT_CONFIG_NAME = f"{_DATASETNAME}_source"
|
105 |
+
|
106 |
+
def _info(self) -> datasets.DatasetInfo:
|
107 |
+
|
108 |
+
if self.config.schema == "source":
|
109 |
+
features = datasets.Features(
|
110 |
+
{
|
111 |
+
"no": datasets.Value("string"),
|
112 |
+
"transformed": datasets.Value("string"),
|
113 |
+
"original-for": datasets.Value("string"),
|
114 |
+
"transformation": datasets.Value("string"),
|
115 |
+
}
|
116 |
+
)
|
117 |
+
|
118 |
+
elif self.config.schema == "nusantara_pairs_multi":
|
119 |
+
features = schemas.pairs_multi_features(self.label_classes)
|
120 |
+
|
121 |
+
else:
|
122 |
+
raise NotImplementedError(f"Schema '{self.config.schema}' is not defined.")
|
123 |
+
|
124 |
+
return datasets.DatasetInfo(
|
125 |
+
description=_DESCRIPTION,
|
126 |
+
features=features,
|
127 |
+
homepage=_HOMEPAGE,
|
128 |
+
license=_LICENSE,
|
129 |
+
citation=_CITATION,
|
130 |
+
)
|
131 |
+
|
132 |
+
def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]:
|
133 |
+
"""Returns SplitGenerators."""
|
134 |
+
|
135 |
+
urls = _URLS[self.config.subset_id]
|
136 |
+
data_paths = dl_manager.download(urls)
|
137 |
+
|
138 |
+
ret = [
|
139 |
+
datasets.SplitGenerator(
|
140 |
+
name=datasets.Split.TRAIN,
|
141 |
+
gen_kwargs={"filepath": data_paths["train"]},
|
142 |
+
)
|
143 |
+
]
|
144 |
+
|
145 |
+
if len(data_paths) > 1:
|
146 |
+
ret.extend(
|
147 |
+
[
|
148 |
+
datasets.SplitGenerator(
|
149 |
+
name=datasets.Split.TEST,
|
150 |
+
gen_kwargs={"filepath": data_paths["test"]},
|
151 |
+
),
|
152 |
+
datasets.SplitGenerator(
|
153 |
+
name=datasets.Split.VALIDATION,
|
154 |
+
gen_kwargs={"filepath": data_paths["dev"]},
|
155 |
+
),
|
156 |
+
]
|
157 |
+
)
|
158 |
+
|
159 |
+
return ret
|
160 |
+
|
161 |
+
def _generate_examples(self, filepath: Path) -> Tuple[int, Dict]:
|
162 |
+
"""Yields examples as (key, example) tuples."""
|
163 |
+
|
164 |
+
with open(filepath, "r", encoding="utf8") as f:
|
165 |
+
dataset = list(map(lambda l: l.rstrip("\r\n").split(","), f))
|
166 |
+
|
167 |
+
_assert = set(map(len, dataset))
|
168 |
+
if _assert != {4}:
|
169 |
+
raise AssertionError(f"Expecting exactly 4 fields (no, transformed, base, label), but found: {_assert}")
|
170 |
+
|
171 |
+
_assert = dataset[0]
|
172 |
+
source_columns = ["no", "transformed", "original-for", "transformation"]
|
173 |
+
if _assert != source_columns:
|
174 |
+
raise AssertionError(f"The expected header is not found. {_assert}")
|
175 |
+
|
176 |
+
dataset = dataset[1:]
|
177 |
+
|
178 |
+
if self.config.schema == "source":
|
179 |
+
for key, ex in enumerate(dataset):
|
180 |
+
yield key, dict(zip(source_columns, ex))
|
181 |
+
|
182 |
+
elif self.config.schema == "nusantara_pairs_multi":
|
183 |
+
for key, ex in enumerate(dataset):
|
184 |
+
yield key, {
|
185 |
+
"id": str(key),
|
186 |
+
"text_1": ex[2],
|
187 |
+
"text_2": ex[1],
|
188 |
+
"label": [ex[3]],
|
189 |
+
}
|