Datasets:
Tasks:
Text Classification
Modalities:
Text
Sub-tasks:
entity-linking-classification
Languages:
English
Size:
< 1K
License:
Create SemEval2018Task7.py
Browse files- SemEval2018Task7.py +273 -0
SemEval2018Task7.py
ADDED
@@ -0,0 +1,273 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# I am trying to understand to the following code. Do not use this for any purpose as I do not support this.
|
2 |
+
# Use the original source from https://huggingface.co/datasets/DFKI-SLT/science_ie/raw/main/science_ie.py
|
3 |
+
|
4 |
+
|
5 |
+
# Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
|
6 |
+
#
|
7 |
+
# Licensed under the Apache License, Version 2.0 (the "License");
|
8 |
+
# you may not use this file except in compliance with the License.
|
9 |
+
# You may obtain a copy of the License at
|
10 |
+
#
|
11 |
+
# http://www.apache.org/licenses/LICENSE-2.0
|
12 |
+
#
|
13 |
+
# Unless required by applicable law or agreed to in writing, software
|
14 |
+
# distributed under the License is distributed on an "AS IS" BASIS,
|
15 |
+
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
16 |
+
# See the License for the specific language governing permissions and
|
17 |
+
# limitations under the License.
|
18 |
+
"""Semeval2018Task7 is a dataset that describes the first task on semantic relation extraction and classification in scientific paper abstracts"""
|
19 |
+
|
20 |
+
|
21 |
+
|
22 |
+
import glob
|
23 |
+
import datasets
|
24 |
+
#from path lib import Path
|
25 |
+
from itertools import permutations
|
26 |
+
from spacy.lang.en import English
|
27 |
+
import xml.dom.minidom
|
28 |
+
import xml.etree.ElementTree as ET
|
29 |
+
|
30 |
+
# Find for instance the citation on arxiv or on the dataset repo/website
|
31 |
+
_CITATION = """\
|
32 |
+
@inproceedings{gabor-etal-2018-semeval,
|
33 |
+
title = "{S}em{E}val-2018 Task 7: Semantic Relation Extraction and Classification in Scientific Papers",
|
34 |
+
author = {G{\'a}bor, Kata and
|
35 |
+
Buscaldi, Davide and
|
36 |
+
Schumann, Anne-Kathrin and
|
37 |
+
QasemiZadeh, Behrang and
|
38 |
+
Zargayouna, Ha{\"\i}fa and
|
39 |
+
Charnois, Thierry},
|
40 |
+
booktitle = "Proceedings of the 12th International Workshop on Semantic Evaluation",
|
41 |
+
month = jun,
|
42 |
+
year = "2018",
|
43 |
+
address = "New Orleans, Louisiana",
|
44 |
+
publisher = "Association for Computational Linguistics",
|
45 |
+
url = "https://aclanthology.org/S18-1111",
|
46 |
+
doi = "10.18653/v1/S18-1111",
|
47 |
+
pages = "679--688",
|
48 |
+
abstract = "This paper describes the first task on semantic relation extraction and classification in scientific paper abstracts at SemEval 2018. The challenge focuses on domain-specific semantic relations and includes three different subtasks. The subtasks were designed so as to compare and quantify the effect of different pre-processing steps on the relation classification results. We expect the task to be relevant for a broad range of researchers working on extracting specialized knowledge from domain corpora, for example but not limited to scientific or bio-medical information extraction. The task attracted a total of 32 participants, with 158 submissions across different scenarios.",
|
49 |
+
}
|
50 |
+
"""
|
51 |
+
|
52 |
+
# You can copy an official description
|
53 |
+
_DESCRIPTION = """\
|
54 |
+
This paper describes the first task on semantic relation extraction and classification in scientific paper abstracts at SemEval 2018. The challenge focuses on domain-specific semantic relations and includes three different subtasks. The subtasks were designed so as to compare and quantify the effect of different pre-processing steps on the relation classification results. We expect the task to be relevant for a broad range of researchers working on extracting specialized knowledge from domain corpora, for example but not limited to scientific or bio-medical information extraction. The task attracted a total of 32 participants, with 158 submissions across different scenarios.
|
55 |
+
"""
|
56 |
+
|
57 |
+
# Add a link to an official homepage for the dataset here
|
58 |
+
_HOMEPAGE = "https://github.com/gkata/SemEval2018Task7/tree/testing"
|
59 |
+
|
60 |
+
# Add the licence for the dataset here if you can find it
|
61 |
+
_LICENSE = ""
|
62 |
+
|
63 |
+
# Add link to the official dataset URLs here
|
64 |
+
# The HuggingFace Datasets library doesn't host the datasets but only points to the original files.
|
65 |
+
# This can be an arbitrary nested dict/list of URLs (see below in `_split_generators` method)
|
66 |
+
_URLS = {
|
67 |
+
"clean": {
|
68 |
+
"train": {
|
69 |
+
"relations": "https://raw.githubusercontent.com/gkata/SemEval2018Task7/testing/1.1.relations.txt",
|
70 |
+
"text": "https://raw.githubusercontent.com/gkata/SemEval2018Task7/testing/1.1.text.xml",
|
71 |
+
},
|
72 |
+
"test": {
|
73 |
+
"relations": "https://raw.githubusercontent.com/gkata/SemEval2018Task7/testing/1.1.test.relations.txt",
|
74 |
+
"text": "https://raw.githubusercontent.com/gkata/SemEval2018Task7/testing/1.1.test.text.xml",
|
75 |
+
},
|
76 |
+
},
|
77 |
+
"noisy": {
|
78 |
+
"train": {
|
79 |
+
"relations": "https://raw.githubusercontent.com/gkata/SemEval2018Task7/testing/1.2.relations.txt",
|
80 |
+
"text": "https://raw.githubusercontent.com/gkata/SemEval2018Task7/testing/1.2.text.xml",
|
81 |
+
},
|
82 |
+
"test": {
|
83 |
+
"relations": "https://raw.githubusercontent.com/gkata/SemEval2018Task7/testing/1.2.test.relations.txt",
|
84 |
+
"text": "https://raw.githubusercontent.com/gkata/SemEval2018Task7/testing/1.2.test.text.xml",
|
85 |
+
},
|
86 |
+
}
|
87 |
+
|
88 |
+
}
|
89 |
+
|
90 |
+
|
91 |
+
def all_text_nodes(root):
|
92 |
+
if root.text is not None:
|
93 |
+
yield root.text
|
94 |
+
for child in root:
|
95 |
+
if child.tail is not None:
|
96 |
+
yield child.tail
|
97 |
+
|
98 |
+
|
99 |
+
def reading_entity_data(string_conver):
|
100 |
+
|
101 |
+
parsing_tag = False
|
102 |
+
final_string = ""
|
103 |
+
tag_string= ""
|
104 |
+
current_tag_id = ""
|
105 |
+
current_tag_starting_pos = 0
|
106 |
+
current_tag_ending_pos= 0
|
107 |
+
entity_mapping_list=[]
|
108 |
+
|
109 |
+
for i in string_conver:
|
110 |
+
if i=='<':
|
111 |
+
parsing_tag = True
|
112 |
+
if current_tag_id!="":
|
113 |
+
current_tag_ending_pos = len(final_string)-1
|
114 |
+
entity_mapping_list.append({"id":current_tag_id,
|
115 |
+
"char_start":current_tag_starting_pos,
|
116 |
+
"char_end":current_tag_ending_pos})
|
117 |
+
current_tag_id= ""
|
118 |
+
tag_string=""
|
119 |
+
|
120 |
+
|
121 |
+
elif i=='>':
|
122 |
+
parsing_tag = False
|
123 |
+
tag_string_split = tag_string.split('"')
|
124 |
+
if len(tag_string_split)>1:
|
125 |
+
current_tag_id= tag_string.split('"')[1]
|
126 |
+
current_tag_starting_pos = len(final_string)
|
127 |
+
|
128 |
+
else:
|
129 |
+
if parsing_tag!=True:
|
130 |
+
final_string = final_string + i
|
131 |
+
else:
|
132 |
+
tag_string = tag_string + i
|
133 |
+
|
134 |
+
return {"abstract":final_string, "entities":entity_mapping_list}
|
135 |
+
|
136 |
+
|
137 |
+
|
138 |
+
|
139 |
+
class Semeval2018Task7(datasets.GeneratorBasedBuilder):
|
140 |
+
"""
|
141 |
+
Semeval2018Task7 is a dataset for semantic relation extraction and classification in scientific paper abstracts
|
142 |
+
"""
|
143 |
+
|
144 |
+
VERSION = datasets.Version("1.1.0")
|
145 |
+
|
146 |
+
BUILDER_CONFIGS = [
|
147 |
+
datasets.BuilderConfig(name="clean", version=VERSION,
|
148 |
+
description="Relation classification on clean data"),
|
149 |
+
datasets.BuilderConfig(name="noisy", version=VERSION,
|
150 |
+
description="Relation classification on noisy data"),
|
151 |
+
]
|
152 |
+
DEFAULT_CONFIG_NAME = "clean"
|
153 |
+
|
154 |
+
def _info(self):
|
155 |
+
class_labels = ["USAGE", "RESULT", "MODEL-FEATURE", "PART_WHOLE", "TOPIC", "COMPARE"]
|
156 |
+
features = datasets.Features(
|
157 |
+
{
|
158 |
+
"id": datasets.Value("string"),
|
159 |
+
"title": datasets.Value("string"),
|
160 |
+
"abstract": datasets.Value("string"),
|
161 |
+
"entities": [
|
162 |
+
{
|
163 |
+
"id": datasets.Value("string"),
|
164 |
+
"char_start": datasets.Value("int32"),
|
165 |
+
"char_end": datasets.Value("int32")
|
166 |
+
}
|
167 |
+
],
|
168 |
+
"relation": [
|
169 |
+
{
|
170 |
+
"label": datasets.ClassLabel(names=class_labels),
|
171 |
+
"arg1": datasets.Value("string"),
|
172 |
+
"arg2": datasets.Value("string"),
|
173 |
+
"reverse": datasets.Value("bool")
|
174 |
+
}
|
175 |
+
]
|
176 |
+
}
|
177 |
+
)
|
178 |
+
|
179 |
+
return datasets.DatasetInfo(
|
180 |
+
# This is the description that will appear on the datasets page.
|
181 |
+
description=_DESCRIPTION,
|
182 |
+
# This defines the different columns of the dataset and their types
|
183 |
+
features=features, # Here we define them above because they are different between the two configurations
|
184 |
+
# If there's a common (input, target) tuple from the features, uncomment supervised_keys line below and
|
185 |
+
# specify them. They'll be used if as_supervised=True in builder.as_dataset.
|
186 |
+
# supervised_keys=("sentence", "label"),
|
187 |
+
# Homepage of the dataset for documentation
|
188 |
+
homepage=_HOMEPAGE,
|
189 |
+
# License for the dataset if available
|
190 |
+
license=_LICENSE,
|
191 |
+
# Citation for the dataset
|
192 |
+
citation=_CITATION,
|
193 |
+
)
|
194 |
+
|
195 |
+
def _split_generators(self, dl_manager):
|
196 |
+
# If several configurations are possible (listed in BUILDER_CONFIGS), the configuration selected by the user is in self.config.name
|
197 |
+
|
198 |
+
# dl_manager is a datasets.download.DownloadManager that can be used to download and extract URLS
|
199 |
+
# It can accept any type or nested list/dict and will give back the same structure with the url replaced with path to local files.
|
200 |
+
# By default the archives will be extracted and a path to a cached folder where they are extracted is returned instead of the archive
|
201 |
+
urls = _URLS[self.config.name]
|
202 |
+
downloaded_files = dl_manager.download(urls)
|
203 |
+
print(downloaded_files)
|
204 |
+
|
205 |
+
return [
|
206 |
+
datasets.SplitGenerator(
|
207 |
+
name=datasets.Split.TRAIN,
|
208 |
+
# These kwargs will be passed to _generate_examples
|
209 |
+
gen_kwargs={
|
210 |
+
"relation_filepath": downloaded_files[datasets.Split.TRAIN]["relations"],
|
211 |
+
"text_filepath": downloaded_files[datasets.Split.TRAIN]["text"],
|
212 |
+
}
|
213 |
+
|
214 |
+
)]
|
215 |
+
# TODO: test split does not contain relations, how to do
|
216 |
+
|
217 |
+
# method parameters are unpacked from `gen_kwargs` as given in `_split_generators`
|
218 |
+
def _generate_examples(self, relation_filepath, text_filepath):
|
219 |
+
# TODO: This method handles input defined in _split_generators to yield (key, example) tuples from the dataset.
|
220 |
+
# The `key` is for legacy reasons (tfds) and is not important in itself, but must be unique for each example.
|
221 |
+
with open(relation_filepath, encoding="utf-8") as f:
|
222 |
+
relations = []
|
223 |
+
|
224 |
+
for key, row in enumerate(f):
|
225 |
+
row_split = row.strip("\n").split("(")
|
226 |
+
use_case = row_split[0]
|
227 |
+
second_half = row_split[1].strip(")")
|
228 |
+
second_half_splits = second_half.split(",")
|
229 |
+
size = len(second_half_splits)
|
230 |
+
|
231 |
+
X = second_half_splits[0]
|
232 |
+
Y = second_half_splits[1]
|
233 |
+
|
234 |
+
relation = {
|
235 |
+
"label": use_case,
|
236 |
+
"arg1": X,
|
237 |
+
"arg2": Y,
|
238 |
+
"reverse": True if size == 3 else False
|
239 |
+
}
|
240 |
+
relations.append(relation)
|
241 |
+
|
242 |
+
doc2 = ET.parse(text_filepath)
|
243 |
+
root = doc2.getroot()
|
244 |
+
|
245 |
+
for child in root:
|
246 |
+
if child.attrib!= None:
|
247 |
+
text_id = child.attrib
|
248 |
+
else:
|
249 |
+
continue
|
250 |
+
|
251 |
+
if child.find("title")!=None:
|
252 |
+
title = child.find("title").text
|
253 |
+
child_abstract = child.find("abstract")
|
254 |
+
else:
|
255 |
+
continue
|
256 |
+
|
257 |
+
if child_abstract!=None:
|
258 |
+
prev=ET.tostring(child_abstract,"utf-8")
|
259 |
+
prev= prev.decode('utf8').replace("b\'","")
|
260 |
+
prev= prev.replace("<abstract>","")
|
261 |
+
prev= prev.replace("</abstract>","")
|
262 |
+
final_list= reading_entity_data(prev)
|
263 |
+
else:
|
264 |
+
continue
|
265 |
+
|
266 |
+
|
267 |
+
yield text_id['id'], {
|
268 |
+
"id": text_id['id'],
|
269 |
+
"title": title,
|
270 |
+
"abstract": final_list['abstract'],
|
271 |
+
"entities": final_list['entities'],
|
272 |
+
"relation": relations
|
273 |
+
}
|