Datasets:

Languages:
French
License:
schlevik commited on
Commit
56aff8b
1 Parent(s): ffb662c

change the bigbio text format to correct text

Browse files
Files changed (1) hide show
  1. essai.py +114 -87
essai.py CHANGED
@@ -128,95 +128,122 @@ class ESSAI(datasets.GeneratorBasedBuilder):
128
 
129
  def _generate_examples(self, datadir):
130
  key = 0
131
- for file in ["ESSAI_neg.txt", "ESSAI_spec.txt"]:
132
- filepath = os.path.join(datadir, file)
133
- label = "negation" if "neg" in file else "speculation"
134
- id_docs = []
135
- id_words = []
136
- words = []
137
- lemmas = []
138
- POS_tags = []
139
-
140
- with open(filepath) as f:
141
- for line in f.readlines():
142
- line_content = line.split("\t")
143
- if len(line_content) > 1:
144
- id_docs.append(line_content[0])
145
- id_words.append(line_content[1])
146
- words.append(line_content[2])
147
- lemmas.append(line_content[3])
148
- POS_tags.append(line_content[4])
149
-
150
- dic = {
151
- "id_docs": np.array(list(map(int, id_docs))),
152
- "id_words": id_words,
153
- "words": words,
154
- "lemmas": lemmas,
155
- "POS_tags": POS_tags,
156
- }
157
- if self.config.schema == "source":
158
- for doc_id in set(dic["id_docs"]):
159
- idces = np.argwhere(dic["id_docs"] == doc_id)[:, 0]
160
- text = [dic["words"][id] for id in idces]
161
- text_lemmas = [dic["lemmas"][id] for id in idces]
162
- POS_tags_ = [dic["POS_tags"][id] for id in idces]
163
- yield key, {
164
- "id": key,
165
- "document_id": doc_id,
166
- "text": text,
167
- "lemmas": text_lemmas,
168
- "POS_tags": POS_tags_,
169
- "labels": [label],
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
170
  }
171
- key += 1
172
- elif self.config.schema == "bigbio_text":
173
- for doc_id in set(dic["id_docs"]):
174
- idces = np.argwhere(dic["id_docs"] == doc_id)[:, 0]
175
- text = " ".join([dic["words"][id] for id in idces])
176
- yield key, {
177
  "id": key,
178
- "document_id": doc_id,
179
- "text": text,
180
- "labels": [label],
181
- }
182
- key += 1
183
- elif self.config.schema == "bigbio_kb":
184
- for doc_id in set(dic["id_docs"]):
185
- idces = np.argwhere(dic["id_docs"] == doc_id)[:, 0]
186
- text = [dic["words"][id] for id in idces]
187
- POS_tags_ = [dic["POS_tags"][id] for id in idces]
188
-
189
- data = {
190
- "id": str(key),
191
- "document_id": doc_id,
192
- "passages": [],
193
- "entities": [],
194
- "relations": [],
195
- "events": [],
196
- "coreferences": [],
197
  }
 
198
  key += 1
199
 
200
- data["passages"] = [
201
- {
202
- "id": str(key + i),
203
- "type": "sentence",
204
- "text": [text[i]],
205
- "offsets": [[i, i + 1]],
206
- }
207
- for i in range(len(text))
208
- ]
209
- key += len(text)
210
-
211
- for i in range(len(text)):
212
- entity = {
213
- "id": key,
214
- "type": "POS_tag",
215
- "text": [POS_tags_[i]],
216
- "offsets": [[i, i + 1]],
217
- "normalized": [],
218
- }
219
- data["entities"].append(entity)
220
- key += 1
221
-
222
- yield key, data
 
128
 
129
  def _generate_examples(self, datadir):
130
  key = 0
131
+ # for file in ["ESSAI_neg.txt", "ESSAI_spec.txt"]:
132
+ filepath = os.path.join(datadir, "ESSAI_neg.txt")
133
+
134
+ filepath2 = os.path.join(datadir, 'ESSAI_spec.txt')
135
+ # label = "negation" if "neg" in file else "speculation"
136
+ id_docs = []
137
+ id_docs_2 = []
138
+ id_words = []
139
+ words = []
140
+ lemmas = []
141
+ POS_tags = []
142
+ NER_tags = []
143
+ NER_tags_2 = []
144
+
145
+ with open(filepath) as f:
146
+ for line in f.readlines():
147
+ line_content = line.split("\t")
148
+ if len(line_content) > 1:
149
+ id_docs.append(line_content[0])
150
+ id_words.append(line_content[1])
151
+ words.append(line_content[2])
152
+ lemmas.append(line_content[3])
153
+ POS_tags.append(line_content[4])
154
+ NER_tags.append(line_content[5].strip())
155
+
156
+ with open(filepath2) as f:
157
+ for line in f.readlines():
158
+ line_content = line.split("\t")
159
+ if len(line_content) > 1:
160
+ id_docs_2.append(line_content[0])
161
+ NER_tags_2.append(line_content[5].strip())
162
+
163
+ dic = {
164
+ "id_docs": np.array(list(map(int, id_docs))),
165
+ "id_words": id_words,
166
+ "words": words,
167
+ "lemmas": lemmas,
168
+ "POS_tags": POS_tags,
169
+ "NER_tags": NER_tags
170
+ }
171
+ dic2 = {
172
+ "id_docs": np.array(list(map(int, id_docs_2))),
173
+ "NER_tags": NER_tags_2
174
+ }
175
+ if self.config.schema == "source":
176
+ for doc_id in set(dic["id_docs"]):
177
+ idces = np.argwhere(dic["id_docs"] == doc_id)[:, 0]
178
+ text = [dic["words"][id] for id in idces]
179
+ text_lemmas = [dic["lemmas"][id] for id in idces]
180
+ POS_tags_ = [dic["POS_tags"][id] for id in idces]
181
+ yield key, {
182
+ "id": key,
183
+ "document_id": doc_id,
184
+ "text": text,
185
+ "lemmas": text_lemmas,
186
+ "POS_tags": POS_tags_,
187
+ "labels": [],
188
+ }
189
+ key += 1
190
+ elif self.config.schema == "bigbio_text":
191
+ for doc_id in set(dic["id_docs"]):
192
+ idces = np.argwhere(dic["id_docs"] == doc_id)[:, 0]
193
+ idces_2 = np.argwhere(dic2["id_docs"] == doc_id)[:, 0]
194
+
195
+ text = " ".join([dic["words"][id] for id in idces])
196
+ label_tokens = [dic["NER_tags"][id] for id in idces]
197
+ label2_tokens = [dic2["NER_tags"][id] for id in idces_2]
198
+ label_ = []
199
+ if not all(l == '***' for l in label_tokens):
200
+ label_.append("negation")
201
+ if not all(l == '***' for l in label2_tokens):
202
+ label_.append("speculation")
203
+ yield key, {
204
+ "id": key,
205
+ "document_id": doc_id,
206
+ "text": text,
207
+ "labels": label_,
208
+ }
209
+ key += 1
210
+ elif self.config.schema == "bigbio_kb":
211
+ for doc_id in set(dic["id_docs"]):
212
+ idces = np.argwhere(dic["id_docs"] == doc_id)[:, 0]
213
+ text = [dic["words"][id] for id in idces]
214
+ POS_tags_ = [dic["POS_tags"][id] for id in idces]
215
+
216
+ data = {
217
+ "id": str(key),
218
+ "document_id": doc_id,
219
+ "passages": [],
220
+ "entities": [],
221
+ "relations": [],
222
+ "events": [],
223
+ "coreferences": [],
224
+ }
225
+ key += 1
226
+
227
+ data["passages"] = [
228
+ {
229
+ "id": str(key + i),
230
+ "type": "sentence",
231
+ "text": [text[i]],
232
+ "offsets": [[i, i + 1]],
233
  }
234
+ for i in range(len(text))
235
+ ]
236
+ key += len(text)
237
+
238
+ for i in range(len(text)):
239
+ entity = {
240
  "id": key,
241
+ "type": "POS_tag",
242
+ "text": [POS_tags_[i]],
243
+ "offsets": [[i, i + 1]],
244
+ "normalized": [],
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
245
  }
246
+ data["entities"].append(entity)
247
  key += 1
248
 
249
+ yield key, data