Datasets:
Tasks:
Text Classification
Modalities:
Text
Sub-tasks:
entity-linking-classification
Languages:
English
Size:
< 1K
License:
Update SemEval2018Task7.py
Browse files- SemEval2018Task7.py +42 -20
SemEval2018Task7.py
CHANGED
@@ -15,7 +15,7 @@
|
|
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
|
19 |
|
20 |
|
21 |
|
@@ -108,8 +108,15 @@ def all_text_nodes(root):
|
|
108 |
yield child.tail
|
109 |
|
110 |
|
111 |
-
def reading_entity_data(
|
112 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
113 |
parsing_tag = False
|
114 |
final_string = ""
|
115 |
tag_string= ""
|
@@ -117,8 +124,8 @@ def reading_entity_data(string_conver):
|
|
117 |
current_tag_starting_pos = 0
|
118 |
current_tag_ending_pos= 0
|
119 |
entity_mapping_list=[]
|
120 |
-
|
121 |
-
for i in
|
122 |
if i=='<':
|
123 |
parsing_tag = True
|
124 |
if current_tag_id!="":
|
@@ -143,8 +150,7 @@ def reading_entity_data(string_conver):
|
|
143 |
else:
|
144 |
tag_string = tag_string + i
|
145 |
|
146 |
-
return {"
|
147 |
-
|
148 |
|
149 |
|
150 |
|
@@ -212,7 +218,7 @@ class Semeval2018Task7(datasets.GeneratorBasedBuilder):
|
|
212 |
# By default the archives will be extracted and a path to a cached folder where they are extracted is returned instead of the archive
|
213 |
urls = _URLS[self.config.name]
|
214 |
downloaded_files = dl_manager.download(urls)
|
215 |
-
print(downloaded_files)
|
216 |
|
217 |
return [
|
218 |
datasets.SplitGenerator(
|
@@ -232,14 +238,14 @@ class Semeval2018Task7(datasets.GeneratorBasedBuilder):
|
|
232 |
# The `key` is for legacy reasons (tfds) and is not important in itself, but must be unique for each example.
|
233 |
with open(relation_filepath, encoding="utf-8") as f:
|
234 |
relations = []
|
235 |
-
|
236 |
for key, row in enumerate(f):
|
237 |
row_split = row.strip("\n").split("(")
|
238 |
use_case = row_split[0]
|
239 |
second_half = row_split[1].strip(")")
|
240 |
second_half_splits = second_half.split(",")
|
241 |
size = len(second_half_splits)
|
242 |
-
|
243 |
relation = {
|
244 |
"label": use_case,
|
245 |
"arg1": second_half_splits[0],
|
@@ -247,6 +253,20 @@ class Semeval2018Task7(datasets.GeneratorBasedBuilder):
|
|
247 |
"reverse": True if size == 3 else False
|
248 |
}
|
249 |
relations.append(relation)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
250 |
|
251 |
doc2 = ET.parse(text_filepath)
|
252 |
root = doc2.getroot()
|
@@ -255,23 +275,25 @@ class Semeval2018Task7(datasets.GeneratorBasedBuilder):
|
|
255 |
if child.find("title")==None:
|
256 |
continue
|
257 |
text_id = child.attrib
|
|
|
258 |
|
259 |
if child.find("abstract")==None:
|
260 |
continue
|
261 |
title = child.find("title").text
|
262 |
child_abstract = child.find("abstract")
|
263 |
|
264 |
-
prev=ET.tostring(child_abstract,"utf-8")
|
265 |
-
prev= prev.decode('utf8').replace("b\'","")
|
266 |
-
prev= prev.replace("<abstract>","")
|
267 |
-
prev= prev.replace("</abstract>","")
|
268 |
-
final_list= reading_entity_data(prev)
|
269 |
|
270 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
271 |
yield text_id['id'], {
|
272 |
"id": text_id['id'],
|
273 |
-
"title":
|
274 |
-
"abstract":
|
275 |
-
"entities":
|
276 |
-
"relation":
|
277 |
}
|
|
|
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 |
|
|
|
108 |
yield child.tail
|
109 |
|
110 |
|
111 |
+
def reading_entity_data(ET_data_to_convert):
|
112 |
+
parsed_data = ET.tostring(ET_data_to_convert,"utf-8")
|
113 |
+
parsed_data= parsed_data.decode('utf8').replace("b\'","")
|
114 |
+
parsed_data= parsed_data.replace("<abstract>","")
|
115 |
+
parsed_data= parsed_data.replace("</abstract>","")
|
116 |
+
parsed_data= parsed_data.replace("<title>","")
|
117 |
+
parsed_data= parsed_data.replace("</title>","")
|
118 |
+
parsed_data = parsed_data.replace("\n\n\n","")
|
119 |
+
|
120 |
parsing_tag = False
|
121 |
final_string = ""
|
122 |
tag_string= ""
|
|
|
124 |
current_tag_starting_pos = 0
|
125 |
current_tag_ending_pos= 0
|
126 |
entity_mapping_list=[]
|
127 |
+
|
128 |
+
for i in parsed_data:
|
129 |
if i=='<':
|
130 |
parsing_tag = True
|
131 |
if current_tag_id!="":
|
|
|
150 |
else:
|
151 |
tag_string = tag_string + i
|
152 |
|
153 |
+
return {"text_data":final_string, "entities":entity_mapping_list}
|
|
|
154 |
|
155 |
|
156 |
|
|
|
218 |
# By default the archives will be extracted and a path to a cached folder where they are extracted is returned instead of the archive
|
219 |
urls = _URLS[self.config.name]
|
220 |
downloaded_files = dl_manager.download(urls)
|
221 |
+
#print(downloaded_files)
|
222 |
|
223 |
return [
|
224 |
datasets.SplitGenerator(
|
|
|
238 |
# The `key` is for legacy reasons (tfds) and is not important in itself, but must be unique for each example.
|
239 |
with open(relation_filepath, encoding="utf-8") as f:
|
240 |
relations = []
|
241 |
+
text_id_to_relations_map= {}
|
242 |
for key, row in enumerate(f):
|
243 |
row_split = row.strip("\n").split("(")
|
244 |
use_case = row_split[0]
|
245 |
second_half = row_split[1].strip(")")
|
246 |
second_half_splits = second_half.split(",")
|
247 |
size = len(second_half_splits)
|
248 |
+
|
249 |
relation = {
|
250 |
"label": use_case,
|
251 |
"arg1": second_half_splits[0],
|
|
|
253 |
"reverse": True if size == 3 else False
|
254 |
}
|
255 |
relations.append(relation)
|
256 |
+
|
257 |
+
arg_id = second_half_splits[0].split(".")[0]
|
258 |
+
if arg_id not in text_id_to_relations_map:
|
259 |
+
text_id_to_relations_map[arg_id] = [relation]
|
260 |
+
else:
|
261 |
+
text_id_to_relations_map[arg_id].append(relation)
|
262 |
+
#print("result", text_id_to_relations_map)
|
263 |
+
|
264 |
+
#for arg_id, values in text_id_to_relations_map.items():
|
265 |
+
#print(f"ID: {arg_id}")
|
266 |
+
# for value in values:
|
267 |
+
# (value)
|
268 |
+
|
269 |
+
|
270 |
|
271 |
doc2 = ET.parse(text_filepath)
|
272 |
root = doc2.getroot()
|
|
|
275 |
if child.find("title")==None:
|
276 |
continue
|
277 |
text_id = child.attrib
|
278 |
+
#print("text_id", text_id)
|
279 |
|
280 |
if child.find("abstract")==None:
|
281 |
continue
|
282 |
title = child.find("title").text
|
283 |
child_abstract = child.find("abstract")
|
284 |
|
|
|
|
|
|
|
|
|
|
|
285 |
|
286 |
+
abstract_text_and_entities = reading_entity_data(child.find("abstract"))
|
287 |
+
title_text_and_entities = reading_entity_data(child.find("title"))
|
288 |
+
|
289 |
+
text_relations = []
|
290 |
+
if text_id['id'] in text_id_to_relations_map:
|
291 |
+
text_relations = text_id_to_relations_map[text_id['id']]
|
292 |
+
|
293 |
yield text_id['id'], {
|
294 |
"id": text_id['id'],
|
295 |
+
"title": title_text_and_entities['text_data'],
|
296 |
+
"abstract": abstract_text_and_entities['text_data'],
|
297 |
+
"entities": abstract_text_and_entities['entities'] + title_text_and_entities['entities'],
|
298 |
+
"relation": text_relations
|
299 |
}
|