holylovenia
commited on
Commit
•
dd41fde
1
Parent(s):
75643e2
Upload men.py with huggingface_hub
Browse files
men.py
ADDED
@@ -0,0 +1,187 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import json
|
2 |
+
import os
|
3 |
+
from pathlib import Path
|
4 |
+
from typing import Dict, List, Tuple
|
5 |
+
|
6 |
+
import datasets
|
7 |
+
|
8 |
+
from seacrowd.utils.configs import SEACrowdConfig
|
9 |
+
from seacrowd.utils.constants import Licenses, Tasks
|
10 |
+
from seacrowd.utils.schemas import kb_features
|
11 |
+
|
12 |
+
_CITATION = """\
|
13 |
+
@misc{chanthran2024malaysian,
|
14 |
+
title={Malaysian English News Decoded: A Linguistic Resource for Named Entity and Relation Extraction},
|
15 |
+
author={Mohan Raj Chanthran and Lay-Ki Soon and Huey Fang Ong and Bhawani Selvaretnam},
|
16 |
+
year={2024},
|
17 |
+
eprint={2402.14521},
|
18 |
+
archivePrefix={arXiv},
|
19 |
+
primaryClass={cs.CL}
|
20 |
+
}
|
21 |
+
"""
|
22 |
+
|
23 |
+
_DATASETNAME = "men"
|
24 |
+
|
25 |
+
_DESCRIPTION = """\
|
26 |
+
The Malaysian English News (MEN) dataset includes 200 Malaysian English news article with human annotated entities and relations (in total 6,061 entities and 3,268 relation instances).
|
27 |
+
Malaysian English combines elements of standard English with Malay, Chinese, and Indian languages. Four human annotators were split into 2 groups, each group annotated 100 news articles
|
28 |
+
and inter-annotator agreement was calculated between 2 or more annotators working on the same task (entity annotation; F1-score 0.82, relation annotation; F1-score 0.51).
|
29 |
+
"""
|
30 |
+
|
31 |
+
_HOMEPAGE = "https://github.com/mohanraj-nlp/MEN-Dataset/tree/main"
|
32 |
+
|
33 |
+
_LANGUAGES = ["eng"]
|
34 |
+
|
35 |
+
_LICENSE = Licenses.MIT.value
|
36 |
+
|
37 |
+
_LOCAL = False
|
38 |
+
|
39 |
+
_URLS = "https://github.com/mohanraj-nlp/MEN-Dataset/archive/refs/heads/main.zip"
|
40 |
+
|
41 |
+
_SUPPORTED_TASKS = [Tasks.RELATION_EXTRACTION, Tasks.NAMED_ENTITY_RECOGNITION]
|
42 |
+
|
43 |
+
_SOURCE_VERSION = "1.0.0"
|
44 |
+
|
45 |
+
_SEACROWD_VERSION = "2024.06.20"
|
46 |
+
|
47 |
+
|
48 |
+
class MENDataset(datasets.GeneratorBasedBuilder):
|
49 |
+
"""The Malaysian English News dataset comprises 200 articles with 6,061 annotated entities and 3,268 relations.
|
50 |
+
Inter-annotator agreement for entity annotation was high (F1-score 0.82), but lower for relation annotation (F1-score 0.51)."""
|
51 |
+
|
52 |
+
SOURCE_VERSION = datasets.Version(_SOURCE_VERSION)
|
53 |
+
SEACROWD_VERSION = datasets.Version(_SEACROWD_VERSION)
|
54 |
+
|
55 |
+
BUILDER_CONFIGS = [
|
56 |
+
SEACrowdConfig(
|
57 |
+
name=f"{_DATASETNAME}_source",
|
58 |
+
version=SOURCE_VERSION,
|
59 |
+
description=f"{_DATASETNAME} source schema",
|
60 |
+
schema="source",
|
61 |
+
subset_id=f"{_DATASETNAME}",
|
62 |
+
),
|
63 |
+
SEACrowdConfig(
|
64 |
+
name=f"{_DATASETNAME}_seacrowd_kb",
|
65 |
+
version=SEACROWD_VERSION,
|
66 |
+
description=f"{_DATASETNAME} SEACrowd schema",
|
67 |
+
schema="seacrowd_kb",
|
68 |
+
subset_id=f"{_DATASETNAME}",
|
69 |
+
),
|
70 |
+
]
|
71 |
+
|
72 |
+
DEFAULT_CONFIG_NAME = f"{_DATASETNAME}_source"
|
73 |
+
|
74 |
+
def _info(self) -> datasets.DatasetInfo:
|
75 |
+
if self.config.schema == "source":
|
76 |
+
features = datasets.Features(
|
77 |
+
{
|
78 |
+
"article": datasets.Value("string"),
|
79 |
+
"entities": datasets.Sequence({"id": datasets.Value("int64"), "label": datasets.Value("string"), "position": {"start": datasets.Value("int32"), "end": datasets.Value("int32")}}),
|
80 |
+
"relations": datasets.Sequence({"id": datasets.Value("string"), "head": datasets.Value("int32"), "tail": datasets.Value("int32"), "relation": datasets.Value("string"), "relation_source": datasets.Value("string")}),
|
81 |
+
}
|
82 |
+
)
|
83 |
+
|
84 |
+
elif self.config.schema == "seacrowd_kb":
|
85 |
+
features = kb_features
|
86 |
+
|
87 |
+
return datasets.DatasetInfo(
|
88 |
+
description=_DESCRIPTION,
|
89 |
+
features=features,
|
90 |
+
homepage=_HOMEPAGE,
|
91 |
+
license=_LICENSE,
|
92 |
+
citation=_CITATION,
|
93 |
+
)
|
94 |
+
|
95 |
+
def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]:
|
96 |
+
"""Returns SplitGenerators."""
|
97 |
+
data_dir = dl_manager.download_and_extract(_URLS)
|
98 |
+
|
99 |
+
return [
|
100 |
+
datasets.SplitGenerator(
|
101 |
+
name=datasets.Split.TRAIN,
|
102 |
+
gen_kwargs={
|
103 |
+
"filepath": data_dir,
|
104 |
+
},
|
105 |
+
),
|
106 |
+
]
|
107 |
+
|
108 |
+
def _MEN_repo_splitter(self, filepath: Path) -> Dict:
|
109 |
+
articles = {}
|
110 |
+
entities = os.path.join(filepath, "MEN-Dataset-main/data/annotated_set.json")
|
111 |
+
relations = os.path.join(filepath, "MEN-Dataset-main/data/rel2id.json")
|
112 |
+
|
113 |
+
with open(entities, "r") as annot_json:
|
114 |
+
annots = json.load(annot_json)
|
115 |
+
|
116 |
+
article_ids = [i["id"] for i in annots]
|
117 |
+
for article_id in article_ids:
|
118 |
+
articles[article_id] = os.path.join(filepath, f"MEN-Dataset-main/data/article_text/{article_id}.txt")
|
119 |
+
|
120 |
+
data_dir = {"entities": entities, "articles": articles, "relations": relations}
|
121 |
+
|
122 |
+
return data_dir
|
123 |
+
|
124 |
+
def _generate_examples(self, filepath: Path) -> Tuple[int, Dict]:
|
125 |
+
"""Yields examples as (key, example) tuples."""
|
126 |
+
filepath = self._MEN_repo_splitter(filepath)
|
127 |
+
|
128 |
+
with open(filepath["entities"], "r") as entities_json:
|
129 |
+
entities = json.load(entities_json)
|
130 |
+
|
131 |
+
articles = {}
|
132 |
+
for article_id in [i["id"] for i in entities]:
|
133 |
+
with open(filepath["articles"][article_id], "r") as article_txt:
|
134 |
+
article = article_txt.read()
|
135 |
+
articles[article_id] = article
|
136 |
+
|
137 |
+
i = 0
|
138 |
+
for item in entities:
|
139 |
+
article_id = item["id"]
|
140 |
+
entities = item["entities"]
|
141 |
+
relations = item["relations"]
|
142 |
+
|
143 |
+
i += 1
|
144 |
+
if self.config.schema == "source":
|
145 |
+
yield i, {
|
146 |
+
"article": articles[article_id],
|
147 |
+
"entities": [
|
148 |
+
{
|
149 |
+
"id": entity["id"],
|
150 |
+
"label": entity["label"],
|
151 |
+
"position": {
|
152 |
+
"start": entity["position"]["start_offset"],
|
153 |
+
"end": entity["position"]["end_offset"],
|
154 |
+
},
|
155 |
+
}
|
156 |
+
for entity in entities
|
157 |
+
],
|
158 |
+
"relations": [{"id": relation["id"], "head": relation["head"], "tail": relation["tail"], "relation": relation["relation"], "relation_source": relation["relation_source"]} for relation in relations],
|
159 |
+
}
|
160 |
+
|
161 |
+
elif self.config.schema == "seacrowd_kb":
|
162 |
+
yield i, {
|
163 |
+
"id": str(i),
|
164 |
+
"passages": [{"id": article_id, "type": "text", "text": [articles[article_id]], "offsets": [[0, len(articles[article_id])]]}],
|
165 |
+
"entities": [
|
166 |
+
{
|
167 |
+
"id": f"{article_id}-entity-{entity['id']}",
|
168 |
+
"type": entity["label"],
|
169 |
+
"text": [articles[article_id][entity["position"]["start_offset"]:entity["position"]["end_offset"]]],
|
170 |
+
"offsets": [[entity["position"]["start_offset"], entity["position"]["end_offset"]]],
|
171 |
+
"normalized": [],
|
172 |
+
}
|
173 |
+
for entity in entities
|
174 |
+
],
|
175 |
+
"events": [],
|
176 |
+
"coreferences": [],
|
177 |
+
"relations": [
|
178 |
+
{
|
179 |
+
"id": f"{article_id}-relation-{relation['id']}",
|
180 |
+
"type": relation["relation"],
|
181 |
+
"arg1_id": f"{article_id}-entity-{relation['head']}",
|
182 |
+
"arg2_id": f"{article_id}-entity-{relation['tail']}",
|
183 |
+
"normalized": [{"db_name": relation["relation_source"], "db_id": ""}],
|
184 |
+
}
|
185 |
+
for relation in relations
|
186 |
+
],
|
187 |
+
}
|