-
Notifications
You must be signed in to change notification settings - Fork 0
/
Function.py
207 lines (162 loc) · 6.58 KB
/
Function.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
# coding:utf-8
import jieba,chardet
import jieba.posseg as pseg
import math
import os
# ================== Function ================
global endMark,compoundWordFlag,filterNounlist,filterVerblist # 也许可以不用声明(默认全局)
endMark = [u'。',u'.',u',',u',',u'\n']
compoundWordFlag = [u'n',u'v',u'a',u'b',u'j',u'l',u'f'] # 添加方位词词性
filterNounlist = [u'n',u'l',u'j',u't']
filterVerblist = [u'v',u'f']
blackFlag = [u'ns']
# ================= TEXT SEG ==============
## change the ltrator to list wit tuple
def SegText(text):
# problem 1: text is not a unicode object?
## getSegResult with compound word, turn to list
result = pseg.cut(text)
textList = []
for word,flag in result:
textList.append((word,flag))
## get compoundUserdict
compoundUserdict = CompoundWord(textList)
jieba.load_userdict(compoundUserdict) #添加组合词后会对分词判断产生影响
## seg text
result = pseg.cut(text)
segTextList = []
for word,flag in result:
segTextList.append((word,flag))
return segTextList
# ================== GET KEYWORD ===============
def ArrangeKeyword(wordDict,keywordNum,weight):
dict = sorted(wordDict.items(),key = lambda item:item[1],reverse = True)
resultList = []
for word,score in dict:
flag = 0
for old_word in resultList:
if CosWord(word,old_word):
flag = 1
if len(word) >= len(old_word):
i = resultList.index(old_word)
resultList[i] = word # 可能会导致权重稍低的词语排序靠前
# resultList = list(set(resultList)) # get the intersection(取不重复的部分,会改变词语顺序,似乎没有用)
if flag: continue
resultList.append(word)
if len(resultList) >= keywordNum:break
# outputResult = ""
# for keyword in resultList:
# if not weight:
# outputResult += "%s\n" % keyword
# else:
# outputResult += "%s %.10f\n" % (keyword,wordDict[keyword])
# outputResult = outputResult[:-1] #除去最后一个换行符
return resultList
def LoadKeywordBlackList(blackwordFilePath):
stopWordList = []
stopFile = open(blackwordFilePath,"rb")
List = stopFile.read()
code = chardet.detect(List)['encoding'] # return the coding UTF8-SIG
unicodeList = str(List,code)
wordList = unicodeList.split(u"\r\n")
for stopword in wordList:
if stopword not in stopWordList:
stopWordList.append(stopword)
return stopWordList
def GetWordPara(segTextList,titleValue = 3.00,nounValue = 1.200,verbValue = 0.800):
TF_Dict = {}
# Now title in the content
global compoundWordFlag,filterNounlist,filterVerblist
cc = True # First line is title
value_Dict = {}
for word,flag in segTextList:
if cc:
if word == u"\n": cc = False
else:
value_Dict[word] = titleValue
if flag[0] in filterNounlist:
if word in value_Dict.keys():
if value_Dict[word] < nounValue:
value_Dict[word] = nounValue
else:
value_Dict[word] = nounValue
elif flag[0] in filterVerblist:
if word not in value_Dict.keys():
value_Dict[word]=float(1.0)
else:
continue
if word not in TF_Dict.keys():TF_Dict[word] = 1
else:TF_Dict[word] += 1
## calculate TF_Value
for word in TF_Dict.keys():
value = TF_Dict[word]
TF_Dict[word] = float(value)/(value+1)
wordValue = {}
for word in TF_Dict.keys():
wordValue[word] = TF_Dict[word] * value_Dict[word]
return wordValue
def CompoundWord(segTextList,compoundT=3):
# set the picture of diagraph by form of dictionary
# 移除停用词列表,使用词性判断
global compoundWordFlag
length = len(segTextList)
wordDigraph = {}
counter = 0-1
while 1:
counter += 1
if counter >= length: break
word,flag = segTextList[counter]
if flag[0] not in compoundWordFlag: continue #只有下一个词符合条件才会开始计算组合词
## 统计组合词出现的次数
tmp_compoundword = [word]
while 1:
counter += 1
if counter >= length: break
next_word,next_flag = segTextList[counter]
if next_flag[0] not in compoundWordFlag: break
tmp_compoundword.append(next_word)
if len(tmp_compoundword) >= 2:
com = ""
for word in tmp_compoundword: com += word
if com not in wordDigraph.keys(): wordDigraph[com] = 1
else: wordDigraph[com] += 1
output = ""
for word in wordDigraph.keys():
if wordDigraph[word] < compoundT:
continue
else:
output += "%s n\n" % word
output += u"%s n\n" % word
output = output[:-1]
tmp_userdict = r"./Dictionary/tmp_compoundword_Dictionary.txt"
openDictFile = open(tmp_userdict,"wb")
openDictFile.write(output.encode('utf8'))
openDictFile.close()
return tmp_userdict
# 计算词语余弦相似度
def CosWord(word1,word2,T=0.7):
wordlist = []
for i in range(0,len(word1)):
if word1[i] not in wordlist:
wordlist.append(word1[i])
for i in range(0,len(word2)):
if word2[i] not in wordlist:
wordlist.append(word2[i])
_vector_list = [[0]*len(wordlist),[0]*len(wordlist)]
for i in range(0,len(word1)):
index = wordlist.index(word1[i])
_vector_list[0][index] = 1
for i in range(0,len(word2)):
index = wordlist.index(word2[i])
_vector_list[1][index] = 1
numerator = 0
denominator = [0,0]
for i in range(0,len(wordlist)):
numerator += _vector_list[0][i] * _vector_list[1][i]
denominator[0] += _vector_list[0][i] ** 2
denominator[1] += _vector_list[1][i] ** 2
denominator[0] = math.sqrt(float(denominator[0]))
denominator[1] = math.sqrt(float(denominator[1]))
result = float(numerator)/(denominator[0] * denominator[1])
if result>T:return True
else:return False