-
Notifications
You must be signed in to change notification settings - Fork 4
/
feature_identification.py
134 lines (114 loc) · 5.14 KB
/
feature_identification.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
"""
The key feature identification tool
"""
import argparse
import subprocess
import csv
import os
from datetime import datetime
from key_feature import KeyFeature
from pathlib import Path
bert_path = 'bert-master'
temp_dir = 'temp'
Path(temp_dir).mkdir(parents=True, exist_ok=True)
def produce_matching_file(feature_path, review_path):
features = []
with open(feature_path, 'r', encoding='utf-8') as file:
for r in file.readlines():
features.append(r.strip())
reviews = []
with open(review_path, 'r', encoding='utf-8') as file:
for r in file.readlines():
reviews.append(r.strip())
# prepare the test data
test_file = temp_dir + '/match_test_set.tsv'
with open(test_file, 'w', encoding='utf-8') as file:
writer = csv.writer(file, delimiter='\t')
for r in reviews:
for f in features:
writer.writerow([0, f, r.split('-*-')[0]])
# do prediction
print('[INFO] running prediction of user review matching ...')
print(' # features = {}, # reviews = {}'.format(len(features), len(reviews)))
subprocess.run('python3 {}/run_classifier.py'
' --task_name=match'
' --do_predict=true'
' --data_dir={}'
' --vocab_file={}/chinese_L-12_H-768_A-12/vocab.txt'
' --bert_config_file={}/chinese_L-12_H-768_A-12/bert_config.json'
' --init_checkpoint={}/model-match'
' --max_seq_length=128'
' --output_dir={}'.format(bert_path, test_file, bert_path, bert_path, bert_path, temp_dir),
shell=True, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
os.remove(temp_dir + '/predict.tf_record')
print('[INFO] prediction result of classifier: {}/test_results.tsv'.format(temp_dir))
predicted = []
with open('{}/test_results.tsv'.format(temp_dir), 'r', encoding='utf-8') as file:
for row in csv.reader(file, delimiter='\t'):
l0, l1 = float(row[0]), float(row[1])
predicted.append('0' if l0 > l1 else '1')
# generate the matching file
result_file = temp_dir + '/match_feature_review.txt'
with open(result_file, 'w', encoding='utf-8') as file:
index = 0
for r in reviews:
for f in features:
file.write(f + '-*-' + r + '-*-' + predicted[index] + '\n')
index += 1
print('[INFO] the matching results between features and user reviews are generated: {}'.format(result_file))
return features, result_file
if __name__ == '__main__':
parser = argparse.ArgumentParser(description='Identify key features of the app based on the given features and user '
'reviews')
parser.add_argument('-f', metavar='FEATURE_FILE', type=str, required=True,
help='the file of specified app features (.txt)')
# the MATCH_FILE should be ordered by date
parser.add_argument('-m', metavar='MATCH_FILE', type=str,
help='the matching results between features and reviews (.txt); if this file is provided, '
'the REVIEW_FILE will be ignored')
parser.add_argument('-r', metavar='REVIEW_FILE', type=str,
help='the file of user reviews (.txt)')
parser.add_argument('-o', metavar='OUTPUT_FILE', type=str, default='key_features.txt',
help='the key features identified (default: key_features.txt)')
parser.add_argument('--bert', type=str, default=bert_path,
help='path of bert directory (default: {})'.format(bert_path))
args = parser.parse_args()
feature_file = args.f
match_file = args.m
review_file = args.r
output_file = args.o
bert_path = args.bert = args.bert
if match_file is None:
# if there is no matching file, generate this file first
features, match_file = produce_matching_file(feature_file, review_file)
else:
features = []
with open(feature_file, 'r', encoding='utf-8') as file:
for line in file.readlines():
features.append(line.strip())
print('[INFO] use the existing user review matching file: {}'.format(match_file))
# get the latest date and date interval
feature_reviews = []
all_dates = []
with open(match_file, 'r', encoding='utf-8') as file:
for line in file.readlines():
feature_reviews.append(line)
line = line.strip().split('-*-')
all_dates.append(datetime.strptime(line[2].split(' ')[0], '%Y-%m-%d'))
all_dates = sorted(all_dates)
earliest_date = all_dates[0]
latest_date = all_dates[-1]
interval = (latest_date - earliest_date).days + 1
print('[INFO] perform analysis with reviews posted between {} and {} ({} days)'.format(
earliest_date.strftime('%Y-%m-%d'), latest_date.strftime('%Y-%m-%d'), interval))
# perform key feature analysis
k = KeyFeature(features, feature_reviews)
_, _, key_features = k.key_feature_identification(latest_date)
feature_str = [e[1] for e in key_features]
print('[INFO] identify {} key features (see {}):'.format(len(key_features), output_file))
print(' {}'.format(feature_str))
with open(output_file, 'w', encoding='utf-8') as file:
for e in feature_str:
file.write(e)
file.write('\n')
print('[TASK FINISHED]')