-
Notifications
You must be signed in to change notification settings - Fork 0
/
framework.py
147 lines (129 loc) · 3.64 KB
/
framework.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
# /usr/bin/python
from features import extract
from time import time
import sys, re
def write_weights(weights, iteration):
filename = "wfinal/"+ str(time()%10000)+"_" + str(iteration) + ".dat"
f = open(filename, "w")
for k,v in weights.iteritems():
f.write(str(k)+" "+str(v)+"\n")
f.close()
sys.stderr.write("weights in " + filename +"\n")
def read_features(featsfile):
featlist = []
feats = open(featsfile, 'r')
while True:
line = feats.readline()
if not line:
break
featlist.append(int(line.strip()))
feats.close()
return featlist
def extract_all_train_features(sents, tagseqs, postagseqs, info):
featset = set([])
i = 0
for sent in sents:
sys.stderr.write(str(i) + "\r")
j = 0
for word in sent:
tag = tagseqs[i][j]
postag = postagseqs[i][j]
if j == 0: # first position
prev = '*'
else:
prev = tagseqs[i][j-1]
featset.update(extract(word, tag, prev, postag, info)) # get a list of all features possible
j += 1
# features for the last label
featset.update(extract('', '<STOP>', tag, '', info))
i += 1
featlist = list(featset)
for f in featlist:
print f
return featlist
def read_data(datafile):
data = open(datafile, 'r')
sents = []
tagseqs = []
postagseqs = []
sent = []
tags = []
postags = []
while 1:
line = data.readline()
if not line:
break
line = line.strip()
if line == "":
sents.append(sent)
tagseqs.append(tags)
postagseqs.append(postags)
sent = []
tags = []
postags = []
continue
word, pos, tag = line.split("\t")
sent.append(word.strip())
tags.append(tag.strip())
postags.append(pos.strip())
data.close()
return sents, tagseqs, postagseqs
def get_maps(sents, postagseqs, gazfile, brownfile):
vocset = set([])
pset = set([])
for sent in sents:
vocset.update(sent)
for postags in postagseqs:
pset.update(postags)
voclist = list(vocset)
plist = list(pset)
i = 1
vocmap = {}
for wtype in voclist:
vocmap[wtype] = i
i += 1
j = 1
pmap = {}
for ptype in plist:
pmap[ptype] = j
j += 1
lmap = {'B':1, 'I':2, 'O':3}
return (vocmap, pmap, lmap, get_gazetteer(gazfile), get_brown(brownfile))
def get_gazetteer(gazfile):
if gazfile == "x.dat":
return None
gmap = {}
g = open(gazfile, 'r')
j = 1
while True:
line = g.readline()
if not line:
break
word, score = line.strip().split(" ")
gmap[word] = int(float(score))
j += 1
g.close()
return gmap
def get_brown(brownfile):
brmap = {}
f = open(brownfile, 'r')
j = 1
while True:
line = f.readline()
if not line:
break
brown, word, num = line.strip().split("\t")
brmap[word] = int(brown)
j += 1
f.close()
return brmap
def get_all(trainfile, gazfile, featfile, brownfile):
sys.stderr.write("reading training data\n")
sents, tagseqs, postagseqs = read_data(trainfile)
info = get_maps(sents, postagseqs, gazfile, brownfile)
sys.stderr.write("extracting features from " + str(len(sents)) + " sentences\n")
#featlist = extract_all_train_features(sents,tagseqs, postagseqs, info)
featlist = read_features(featfile)
return sents, tagseqs, postagseqs, featlist, info
if __name__ == "__main__":
sentset, labelset, postagset, all_feats, info = get_all(sys.argv[1], sys.argv[2], sys.argv[3], sys.argv[4])