-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathutils_ner_bio.py
295 lines (268 loc) · 12.6 KB
/
utils_ner_bio.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
# coding=utf-8
# Copyright 2018 The Google AI Language Team Authors and The HuggingFace Inc. team.
# Copyright (c) 2018, NVIDIA CORPORATION. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
""" Named entity recognition fine-tuning: utilities to work with CoNLL-2003 task. """
import logging
import os
import json
from utils import get_labels
logger = logging.getLogger(__name__)
class InputExample(object):
"""A single training/test example for token classification."""
def __init__(self, id, words, labels):
"""Constructs a InputExample.
Args:
id: Unique id for the example.
words: list. The words of the sequence.
labels: (Optional) list. The labels for each word of the sequence. This should be
specified for train and dev examples, but not for test examples.
"""
self.id = id
self.words = words
self.labels = labels
class InputFeatures(object):
"""A single set of features of data."""
def __init__(self, input_ids, attention_mask, token_type_ids, label_ids):
self.input_ids = input_ids
self.attention_mask = attention_mask
self.token_type_ids = token_type_ids
self.label_ids = label_ids
## lic 格式
def trigger_process_bio_lic(input_file, is_predict=False):
rows = open(input_file, encoding='utf-8').read().splitlines()
results = []
for row in rows:
if len(row)==1: print(row)
row = json.loads(row)
labels = ['O']*len(row["text"])
if is_predict:
results.append({"id":row["id"], "words":list(row["text"]), "labels":labels})
continue
for event in row["event_list"]:
trigger = event["trigger"]
event_type = event["event_type"]
trigger_start_index = event["trigger_start_index"]
labels[trigger_start_index]= "B-{}".format(event_type)
for i in range(1, len(trigger)):
labels[trigger_start_index+i]= "I-{}".format(event_type)
# labels[trigger_start_index+i]= "I-{}".format("触发词")
results.append({"id":row["id"], "words":list(row["text"]), "labels":labels})
# write_file(results,output_file)
return results
## ccks格式
def trigger_process_bio_ccks(input_file, is_predict=False):
rows = open(input_file, encoding='utf-8').read().splitlines()
results = []
for row in rows:
if len(row)==1: print(row)
row = json.loads(row)
labels = ['O']*len(row["content"])
if is_predict:
results.append({"id":row["id"], "words":list(row["content"]), "labels":labels})
continue
for event in row["events"]:
event_type = event["type"]
for mention in event["mentions"]:
if mention["role"]=="trigger":
trigger = mention["word"]
trigger_start_index, trigger_end_index = mention["span"]
labels[trigger_start_index]= "B-{}".format(event_type)
for i in range(trigger_start_index+1, trigger_end_index):
labels[i]= "I-{}".format(event_type)
# labels[i]= "I-{}".format("触发词")
break
results.append({"id":row["id"], "words":list(row["content"]), "labels":labels})
# write_file(results,output_file)
return results
## lic格式
def role_process_bio_lic(input_file, add_event_type_to_role=False, is_predict=False):
rows = open(input_file, encoding='utf-8').read().splitlines()
results = []
for row in rows:
if len(row)==1: print(row)
row = json.loads(row)
labels = ['O']*len(row["text"])
if is_predict:
results.append({"id":row["id"], "words":list(row["text"]), "labels":labels})
continue
for event in row["event_list"]:
# print(event)
event_type = event["event_type"]
for arg in event["arguments"]:
role = arg['role']
if add_event_type_to_role: role = event_type + '-' + role
argument = arg['argument']
argument_start_index = arg["argument_start_index"]
labels[argument_start_index]= "B-{}".format(role)
for i in range(1, len(argument)):
labels[argument_start_index+i]= "I-{}".format(role)
# if arg['alias']!=[]: print(arg['alias'])
results.append({"id":row["id"], "words":list(row["text"]), "labels":labels})
# write_file(results,output_file)
return results
## ccks格式
def role_process_bio_ccks(input_file, add_event_type_to_role=False, is_predict=False):
rows = open(input_file, encoding='utf-8').read().splitlines()
results = []
for row in rows:
if len(row)==1: print(row)
row = json.loads(row)
labels = ['O']*len(row["content"])
if is_predict:
results.append({"id":row["id"], "words":list(row["content"]), "labels":labels})
continue
for event in row["events"]:
event_type = event["type"]
for arg in event["mentions"]:
role = arg['role']
if role=="trigger": continue
if add_event_type_to_role: role = event_type + '-' + role
argument_start_index, argument_end_index = arg["span"]
labels[argument_start_index]= "B-{}".format(role)
for i in range(argument_start_index+1, argument_end_index):
labels[i]= "I-{}".format(role)
# if arg['alias']!=[]: print(arg['alias'])
results.append({"id":row["id"], "words":list(row["content"]), "labels":labels})
# write_file(results,output_file)
return results
def read_examples_from_file(data_dir, mode, task, dataset="ccks"):
file_path = os.path.join(data_dir, "{}.json".format(mode))
if dataset=="ccks":
if task=='trigger': items = trigger_process_bio_ccks(file_path)
elif task=='role': items = role_process_bio_ccks(file_path, add_event_type_to_role=True)
elif dataset=="lic":
if task=='trigger': items = trigger_process_bio_lic(file_path)
elif task=='role': items = role_process_bio_lic(file_path, add_event_type_to_role=True)
return [InputExample(**item) for item in items]
def convert_examples_to_features(
examples,
label_list,
max_seq_length,
tokenizer,
cls_token_at_end=False,
cls_token="[CLS]",
cls_token_segment_id=1,
sep_token="[SEP]",
sep_token_extra=False,
pad_on_left=False,
pad_token=0,
pad_token_segment_id=0,
pad_token_label_id=-100,
sequence_a_segment_id=0,
mask_padding_with_zero=True,
):
""" Loads a data file into a list of `InputBatch`s
`cls_token_at_end` define the location of the CLS token:
- False (Default, BERT/XLM pattern): [CLS] + A + [SEP] + B + [SEP]
- True (XLNet/GPT pattern): A + [SEP] + B + [SEP] + [CLS]
`cls_token_segment_id` define the segment id associated to the CLS token (0 for BERT, 2 for XLNet)
"""
label_map = {label: i for i, label in enumerate(label_list)}
# print(label_map)
features = []
for (ex_index, example) in enumerate(examples):
if ex_index % 10000 == 0:
logger.info("Writing example %d of %d", ex_index, len(examples))
# print(example.words, example.labels)
# print(len(example.words), len(example.labels))
tokens = []
label_ids = []
for word, label in zip(example.words, example.labels):
word_tokens = tokenizer.tokenize(word)
if len(word_tokens)==1:
tokens.extend(word_tokens)
if len(word_tokens)>1:
print(word,">1")
tokens.extend(word_tokens[:1])
if len(word_tokens)<1:
# print(word,"<1") # 基本都是空格
tokens.extend(["[unused1]"])
label_ids.extend([label_map[label]])
# if len(tokens)!= len(label_ids):
# print(word, word_tokens, tokens, label_ids)
assert len(tokens) == len(label_ids)
# print(len(tokens),len(label_ids))
# Account for [CLS] and [SEP] with "- 2" and with "- 3" for RoBERTa.
special_tokens_count = 3 if sep_token_extra else 2
if len(tokens) > max_seq_length - special_tokens_count:
tokens = tokens[: (max_seq_length - special_tokens_count)]
label_ids = label_ids[: (max_seq_length - special_tokens_count)]
# The convention in BERT is:
# (a) For sequence pairs:
# tokens: [CLS] is this jack ##son ##ville ? [SEP] no it is not . [SEP]
# type_ids: 0 0 0 0 0 0 0 0 1 1 1 1 1 1
# (b) For single sequences:
# tokens: [CLS] the dog is hairy . [SEP]
# type_ids: 0 0 0 0 0 0 0
#
# Where "type_ids" are used to indicate whether this is the first
# sequence or the second sequence. The embedding vectors for `type=0` and
# `type=1` were learned during pre-training and are added to the wordpiece
# embedding vector (and position vector). This is not *strictly* necessary
# since the [SEP] token unambiguously separates the sequences, but it makes
# it easier for the model to learn the concept of sequences.
#
# For classification tasks, the first vector (corresponding to [CLS]) is
# used as as the "sentence vector". Note that this only makes sense because
# the entire model is fine-tuned.
tokens += [sep_token]
label_ids += [pad_token_label_id]
if sep_token_extra:
# roberta uses an extra separator b/w pairs of sentences
tokens += [sep_token]
label_ids += [pad_token_label_id]
token_type_ids = [sequence_a_segment_id] * len(tokens)
if cls_token_at_end:
tokens += [cls_token]
label_ids += [pad_token_label_id]
token_type_ids += [cls_token_segment_id]
else:
tokens = [cls_token] + tokens
label_ids = [pad_token_label_id] + label_ids
token_type_ids = [cls_token_segment_id] + token_type_ids
input_ids = tokenizer.convert_tokens_to_ids(tokens)
# print(len(tokens), len(input_ids), len(label_ids))
# The mask has 1 for real tokens and 0 for padding tokens. Only real
# tokens are attended to.
attention_mask = [1 if mask_padding_with_zero else 0] * len(input_ids)
# Zero-pad up to the sequence length.
padding_length = max_seq_length - len(input_ids)
if pad_on_left:
input_ids = ([pad_token] * padding_length) + input_ids
attention_mask = ([0 if mask_padding_with_zero else 1] * padding_length) + attention_mask
token_type_ids = ([pad_token_segment_id] * padding_length) + token_type_ids
label_ids = ([pad_token_label_id] * padding_length) + label_ids
else:
input_ids += [pad_token] * padding_length
attention_mask += [0 if mask_padding_with_zero else 1] * padding_length
token_type_ids += [pad_token_segment_id] * padding_length
label_ids += [pad_token_label_id] * padding_length
# print(len(label_ids), max_seq_length)
assert len(input_ids) == max_seq_length
assert len(attention_mask) == max_seq_length
assert len(token_type_ids) == max_seq_length
assert len(label_ids) == max_seq_length
if ex_index < 5:
logger.info("*** Example ***")
logger.info("id: %s", example.id)
logger.info("tokens: %s", " ".join([str(x) for x in tokens]))
logger.info("input_ids: %s", " ".join([str(x) for x in input_ids]))
logger.info("attention_mask: %s", " ".join([str(x) for x in attention_mask]))
logger.info("token_type_ids: %s", " ".join([str(x) for x in token_type_ids]))
logger.info("label_ids: %s", " ".join([str(x) for x in label_ids]))
features.append(
InputFeatures(input_ids=input_ids, attention_mask=attention_mask, token_type_ids=token_type_ids, label_ids=label_ids)
)
return features