-
Notifications
You must be signed in to change notification settings - Fork 26
/
pipeline.py
173 lines (159 loc) · 7.2 KB
/
pipeline.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
import os
import argparse
import random
import string
import pickle
import gensim
import numpy as np
def cos( vector1, vector2):
return float(np.sum(vector1*vector2))/(np.linalg.norm(vector1)*np.linalg.norm(vector2))
def preprocess_log(ipath, opath):
#preprocess
processed_log = os.path.join(opath, 'without_variables.log')
command_for_preprocessing = "python code/preprocessing.py -rawlog %s -o %s"%(ipath, processed_log)
os.system(command_for_preprocessing)
return processed_log
def generate_oov(processed_log, opath):
generate_file_path = os.path.join(opath, 'changed_log')
if not os.path.exists(generate_file_path):
os.mkdir(generate_file_path)
changed_log = os.path.join(generate_file_path, "without_variables.log")
new_vocab = os.path.join(generate_file_path, "vocab.txt")
old_to_new_dict = os.path.join(generate_file_path, "old_new_dict.txt")
with open(processed_log, 'r') as file:
logs = file.readlines()
temp_result = []
new_words = set()
old_new_dict = {}
for log in logs:
log_in_word = log.split()
log_length = len(log_in_word)
target_word = random.randint(0,log_length-1)
if log_in_word[target_word] in old_to_new_dict:
target_word = (target_word + 1) % log_length
word_length = len(log_in_word[target_word])
target_letter = random.randint(0, word_length-1)
change_letter = random.randint(0, 25)
if log_in_word[target_word][target_letter].lower() == string.ascii_lowercase[change_letter]:
change_letter = (change_letter + 1) % 26
old = log_in_word[target_word]
log_in_word[target_word] = (log_in_word[target_word][:target_letter]
+ string.ascii_lowercase[change_letter]
+ log_in_word[target_word][target_letter+1:] )
new_words.add(log_in_word[target_word]+'\n')
if old in old_new_dict:
old_new_dict[old].append(log_in_word[target_word])
else:
old_new_dict[old] = [log_in_word[target_word]]
temp_result.append(' '.join(log_in_word)+'\n')
with open(changed_log, 'w') as file:
file.writelines(temp_result)
with open(new_vocab, 'w') as file:
file.writelines(new_words)
with open(old_to_new_dict, 'wb') as file:
for key in old_new_dict:
old_new_dict[key] = list(set(old_new_dict[key]))
file.write(pickle.dumps(old_new_dict))
return new_vocab, old_to_new_dict
def pipeline(processed_log, new_vocab):
# Antonyms&Synonyms Extraction
sys_output = os.path.join(opath, 'sys.txt')
ants_output = os.path.join(opath, 'ants.txt')
command_for_AS_extraction = '''python code/get_syn_ant.py -logs %s -ant_file %s -syn_file %s'''%(processed_log, ants_output, sys_output)
os.system(command_for_AS_extraction)
# Relation Triple Extraction
triplet_log = 'triples.txt'
command_for_triplet = '''python code/get_triplet.py %s %s'''%(processed_log, triplet_log)
os.system(command_for_triplet)
# Semantic Word Embedding
train_log = os.path.join(opath, 'for_training.log')
command_for_train = "python code/getTempLogs.py -input %s -output %s" %(processed_log, train_log)
print(command_for_train)
os.system(command_for_train)
# Semantic Word Embedding
train_model = os.path.join(opath, 'embedding.model')
vocab = os.path.join(opath, 'embedding.vocab')
command_for_model = ('''code/LRWE/src/lrcwe -train %s -synonym %s -antonym %s -output %s -save-vocab %s -belta-rel 0.8 -alpha-rel 0.01 -alpha-ant 0.3 -size 32 -min-count 1 -window 2 -triplet %s'''
%(train_log, sys_output,
ants_output, train_model,
vocab, triplet_log))
os.system(command_for_model)
print('------')
print(command_for_model)
oov_words = os.path.join(opath, 'words.pkl')
command_for_oov = "python code/mimick/make_dataset.py --vectors %s --w2v-format --output %s"%(train_model, oov_words)
os.system(command_for_oov)
print('------')
print(command_for_oov)
oov_vector = os.path.join(opath, 'oov.vector')
learning_rate = 0.006
epoch = 20
num_of_layers = 1
dropout = -1
hidden_dim = 250
ch_dim = 36
command_for_new_embedding = ("python code/mimick/model.py --dataset %s --vocab %s --output %s --num-epochs %d --learning-rate %f --num-lstm-layers %d --cosine --dropout %f --all-from-mimick --hidden-dim %d --char-dim %d"
%(oov_words, new_vocab, oov_vector, epoch, learning_rate, num_of_layers, dropout, hidden_dim, ch_dim))
os.system(command_for_new_embedding)
print('------')
print(command_for_new_embedding)
# get log2vec
log_vector = os.path.join(opath, 'log.vector')
command_for_log2vec = " python code/Log2Vec.py -logs %s -word_model %s -log_vector_file %s -dimension 32"%(processed_log, train_model, log_vector)
os.system(command_for_log2vec)
print('------')
print(command_for_log2vec)
return train_model, oov_vector
def evaluate(word_model_path, old_to_new_dict, oov_vector_path, opath):
word2vec = gensim.models.KeyedVectors.load_word2vec_format(word_model_path, binary = False)
oov_vec = gensim.models.KeyedVectors.load_word2vec_format(oov_vector_path, binary = False)
with open(old_to_new_dict, 'rb') as file:
old_new_dict = pickle.loads(file.read())
total_score = 0
count = 0
result = []
for key in old_new_dict:
for new_word in old_new_dict[key]:
if key not in word2vec:
print(key)
continue
score = cos(word2vec[key], oov_vec[new_word])
total_score += score
count += 1
result.append((key, new_word, score))
score_path = os.path.join(opath, 'score')
if not os.path.exists(score_path):
os.mkdir(score_path)
with open(os.path.join(score_path, 'score'), 'w') as ofile:
ofile.write('score: '+str(total_score/count)+'\n')
with open(os.path.join(score_path, 'result.txt'), 'w') as ofile:
for i in result:
ofile.write(i[0]+ ' '+i[1]+' '+str(i[2])+'\n')
return total_score/count, result
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('-i', help='input_file')
parser.add_argument('-o', help='output directory', type=str, default=None)
parser.add_argument('-t', help='log type')
args = parser.parse_args()
ipath = args.i
ipath = os.path.abspath(ipath)
if args.o == None:
output_path = 'oov_result/'
if not os.path.exists(output_path):
os.mkdir('oov_result')
else:
output_path = args.o
if not os.path.exists(output_path):
os.mkdir(output_path)
output_path = os.path.abspath(output_path)
log_type = args.t
opath = os.path.join(output_path, log_type)
if not os.path.exists(opath):
os.mkdir(opath)
processed_log = preprocess_log(ipath, opath)
new_vocab, old_to_new_dict = generate_oov(processed_log, opath)
train_model, oov_vector = pipeline(processed_log, new_vocab)
score, result = evaluate(train_model, old_to_new_dict, oov_vector, opath)
print('---------')
print(score)