-
Notifications
You must be signed in to change notification settings - Fork 394
/
utils.py
204 lines (186 loc) · 7.48 KB
/
utils.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
# coding=utf-8
import codecs
import re
import numpy as np
def calculate(x,y,id2word,id2tag,res=[]):
entity=[]
for i in range(len(x)): #for every sen
for j in range(len(x[0])): #for every word
if x[i][j]==0 or y[i][j]==0:
continue
if id2tag[y[i][j]][0]=='B':
entity=[id2word[x[i][j]]+'/'+id2tag[y[i][j]]]
elif id2tag[y[i][j]][0]=='M' and len(entity)!=0 and entity[-1].split('/')[1][1:]==id2tag[y[i][j]][1:]:
entity.append(id2word[x[i][j]]+'/'+id2tag[y[i][j]])
elif id2tag[y[i][j]][0]=='E' and len(entity)!=0 and entity[-1].split('/')[1][1:]==id2tag[y[i][j]][1:]:
entity.append(id2word[x[i][j]]+'/'+id2tag[y[i][j]])
entity.append(str(i))
entity.append(str(j))
res.append(entity)
entity=[]
else:
entity=[]
return res
def get_entity(x,y,id2tag):
entity=""
res=[]
for i in range(len(x)): #for every sen
for j in range(len(x[0])): #for every word
if y[i][j]==0:
continue
if id2tag[y[i][j]][0]=='B':
entity=id2tag[y[i][j]][1:]+':'+x[i][j]
elif id2tag[y[i][j]][0]=='M' and len(entity)!=0 :
entity+=x[i][j]
elif id2tag[y[i][j]][0]=='E' and len(entity)!=0 :
entity+=x[i][j]
res.append(entity)
entity=[]
else:
entity=[]
return res
def write_entity(outp,x,y,id2tag):
'''
注意，这个函数每次使用是在文档的最后添加新信息。
'''
entity=''
for i in range(len(x)):
if y[i]==0:
continue
if id2tag[y[i]][0]=='B':
entity=id2tag[y[i]][2:]+':'+x[i]
elif id2tag[y[i]][0]=='M' and len(entity)!=0:
entity+=x[i]
elif id2tag[y[i]][0]=='E' and len(entity)!=0:
entity+=x[i]
outp.write(entity+' ')
entity=''
else:
entity=''
return
def train(model,sess,saver,epochs,batch_size,data_train,data_test,id2word,id2tag):
batch_num = int(data_train.y.shape[0] / batch_size)
batch_num_test = int(data_test.y.shape[0] / batch_size)
for epoch in range(epochs):
for batch in range(batch_num):
x_batch, y_batch = data_train.next_batch(batch_size)
# print x_batch.shape
feed_dict = {model.input_data:x_batch, model.labels:y_batch}
pre,_ = sess.run([model.viterbi_sequence,model.train_op], feed_dict)
acc = 0
if batch%200==0:
for i in range(len(y_batch)):
for j in range(len(y_batch[0])):
if y_batch[i][j]==pre[i][j]:
acc+=1
print float(acc)/(len(y_batch)*len(y_batch[0]))
path_name = "./model/model"+str(epoch)+".ckpt"
print path_name
if epoch%3==0:
saver.save(sess, path_name)
print "model has been saved"
entityres=[]
entityall=[]
for batch in range(batch_num):
x_batch, y_batch = data_train.next_batch(batch_size)
feed_dict = {model.input_data:x_batch, model.labels:y_batch}
pre = sess.run([model.viterbi_sequence], feed_dict)
pre = pre[0]
entityres = calculate(x_batch,pre,id2word,id2tag,entityres)
entityall = calculate(x_batch,y_batch,id2word,id2tag,entityall)
jiaoji = [i for i in entityres if i in entityall]
if len(jiaoji)!=0:
zhun = float(len(jiaoji))/len(entityres)
zhao = float(len(jiaoji))/len(entityall)
print "train"
print "zhun:", zhun
print "zhao:", zhao
print "f:", (2*zhun*zhao)/(zhun+zhao)
else:
print "zhun:",0
entityres=[]
entityall=[]
for batch in range(batch_num_test):
x_batch, y_batch = data_test.next_batch(batch_size)
feed_dict = {model.input_data:x_batch, model.labels:y_batch}
pre = sess.run([model.viterbi_sequence], feed_dict)
pre = pre[0]
entityres = calculate(x_batch,pre,id2word,id2tag,entityres)
entityall = calculate(x_batch,y_batch,id2word,id2tag,entityall)
jiaoji = [i for i in entityres if i in entityall]
if len(jiaoji)!=0:
zhun = float(len(jiaoji))/len(entityres)
zhao = float(len(jiaoji))/len(entityall)
print "test"
print "zhun:", zhun
print "zhao:", zhao
print "f:", (2*zhun*zhao)/(zhun+zhao)
else:
print "zhun:",0
max_len = 60
def padding(ids):
if len(ids) >= max_len:
return ids[:max_len]
else:
ids.extend([0]*(max_len-len(ids)))
return ids
def padding_word(sen):
if len(sen) >= max_len:
return sen[:max_len]
else:
return sen
def test_input(model,sess,word2id,id2tag,batch_size):
while True:
text = raw_input("Enter your input: ").decode('utf-8');
text = re.split(u'[，。！？、‘’“”（）]', text)
text_id=[]
for sen in text:
word_id=[]
for word in sen:
if word in word2id:
word_id.append(word2id[word])
else:
word_id.append(word2id["unknow"])
text_id.append(padding(word_id))
zero_padding=[]
zero_padding.extend([0]*max_len)
text_id.extend([zero_padding]*(batch_size-len(text_id)))
feed_dict = {model.input_data:text_id}
pre = sess.run([model.viterbi_sequence], feed_dict)
entity = get_entity(text,pre[0],id2tag)
print 'result:'
for i in entity:
print i
def extraction(input_path,output_path,model,sess,word2id,id2tag,batch_size):
text_id=[]
text = []
with codecs.open(input_path,'rb','utf8') as inp:
for line in inp.readlines():
line = re.split('[，。！？、‘’“”（）]'.decode('utf-8'), line.strip())
for sentence in line:
if sentence=='' or sentence==' ':
continue
word_id=[]
for word in sentence:
if word in word2id:
word_id.append(word2id[word])
else:
word_id.append(word2id["unknow"])
text_id.append(padding(word_id))
text.append(padding_word(sentence))
zero_padding=[]
zero_padding.extend([0]*max_len)
text_id.extend([zero_padding]*(batch_size-len(text_id)%batch_size))
text_id = np.asarray(text_id)
text_id = text_id.reshape(-1,batch_size,max_len)
predict = []
for index in range(len(text_id)):
feed_dict = {model.input_data:text_id[index]}
pre = sess.run([model.viterbi_sequence], feed_dict)
predict.append(pre[0])
predict = np.asarray(predict).reshape(-1,max_len)
with codecs.open(output_path,'a','utf-8') as outp:
for index in range(len(text)):
outp.write(text[index]+" ")
write_entity(outp,text[index],predict[index],id2tag)
outp.write('\n')