def createVocabList():
global mailLists
global mailIndex
global spamPro
global ordPro
global spamNum
global ordNum
spamPro = 0
ordPro = 0
spamDict = {}
ordDict = {}
spamNum = 0
ordNum = 0
for n in range(len(mailLists)):
wordInMail = mecab.MeCabClass(mailLists[n])
NounWords = wordInMail.GetNoun2()
AdjWords = wordInMail.GetAdj()
if mailIndex[n] == 1:
for noun in NounWords:
ordNum = ordNum + 1
addWordToDict(noun,ordDict)
for adj in AdjWords:
ordNum = ordNum + 1
addWordToDict(adj,ordDict)
if mailIndex[n] == 0:
for noun in NounWords:
spamNum = spamNum + 1
addWordToDict(noun,spamDict)
for adj in AdjWords:
spamNum = spamNum + 1
addWordToDict(adj,spamDict)
spamPro = float(spamNum/(spamNum + ordNum))
ordPro = float(ordNum/(spamNum + ordNum))
return spamDict,ordDict
def speechStats1():
global TwoCooccurFile
global ThreeCooccurFile
# TwoWordList = Tools.content_wordList(TwoCooccurFile,':')
TwoWordList = Tools.content_wordList(ThreeCooccurFile, ':')
allSpeech = set()
allSpeechTuple = {}
for word in TwoWordList:
thisWordMec = mecab.MeCabClass(word)
thisSpeechs = thisWordMec.GetAll3()
try:
thisSpeech = thisSpeechs[0]
for n in range(len(thisSpeechs) - 1):
thisSpeech += "+" + thisSpeechs[n + 1]
allSpeech.add(thisSpeech)
except:
pass
allSpeechTuple.setdefault(thisSpeech, []).append(word)
SpeechCntTuple = {}
for speech in allSpeechTuple.keys():
SpeechCntTuple[speech] = len(allSpeechTuple[speech])
sortedList = Tools.TupleSort(SpeechCntTuple)
Speech = []
SpeechCnt = []
Cnt = 0
for ele in sortedList:
Cnt += 1
print Cnt, ':', ele[0], ':\t', ele[1]
Speech.append(Cnt)
SpeechCnt.append(ele[1])
Tools.plot(Speech, SpeechCnt)
def DuplicateRemoval(filename):
fopen = open(filename, 'r')
content = fopen.read()
oneWordList = content.split('\n')
allSpeech = set()
allSpeechTuple = {}
wordTuple = {}
for oneWord in oneWordList:
# thisWord = oneWord.split(':')[0]
thisWord = oneWord
try:
if thisWord in wordTuple.keys():
wordTuple[thisWord] += int(
oneWord.split(':')[1].split(' ')[-1])
else:
wordTuple[thisWord] = int(oneWord.split(':')[1].split(' ')[-1])
except:
traceback.print_exception()
newfilename = filename[:-4] + "_DuplicateRemoval.txt"
newfile = open(newfilename, 'wb')
wordList = Tools.TupleSort(wordTuple)
for word in wordList:
# print word
thisWordMec = mecab.MeCabClass(word[0])
thisSpeechs = thisWordMec.GetAll()
newfile.write(word[0] + ':')
for subword in thisSpeechs:
newfile.write(subword + ' ')
newfile.write(str(wordTuple[word[0]]) + '\n')
newfile.close()
def wordCounter(self):
if self.counter != 0:
return self.counter
else:
for file in self.fileLists:
fopen = open(file, 'r')
content = fopen.read()
nounInCon = mecab.MeCabClass(content).GetNoun2()
self.counter += len(nounInCon)
return self.counter
def speech(filename, newfile, threshold):
oneWordList = Tools.content_wordList(filename, ':')
wordListFre = Tools.content_scoreList(filename, [':', ' '])
# print len(oneWordList),len(WordsFreList)
allSpeech = set()
allSpeechTuple = {}
SpeechWordsFreTup = {}
for n in range(len(oneWordList)):
# try:
# thisWord = oneWord.split(':')[0]
thisWord = oneWordList[n]
thisWordMec = mecab.MeCabClass(thisWord)
thisSpeechs = thisWordMec.GetAll3()
try:
thisSpeech = thisSpeechs[0]
for x in range(len(thisSpeechs) - 1):
thisSpeech += "+" + thisSpeechs[x + 1]
allSpeech.add(thisSpeech)
except:
pass
# if thisWord not in allSpeechTuple[thisSpeech]:
allSpeechTuple.setdefault(thisSpeech, []).append(thisWord)
# try:
if thisSpeech in SpeechWordsFreTup.keys():
SpeechWordsFreTup[thisSpeech] += wordListFre[n]
else:
SpeechWordsFreTup[thisSpeech] = wordListFre[n]
newfile = open(newfile, 'wb')
for speech in allSpeechTuple.keys():
if SpeechWordsFreTup[speech] > threshold:
newfile.write(speech + ':' + str(SpeechWordsFreTup[speech]) +
':\n')
if SpeechWordsFreTup[speech] > 300:
for n in range(300):
try:
x = random.randint(0, len(allSpeechTuple) - 1)
newfile.write(allSpeechTuple[speech][x] + ' ')
except:
pass
continue
else:
for word in allSpeechTuple[speech]:
newfile.write(word + ' ')
newfile.write('\n\n')
newfile.close()
# newfile2 = open('allSpeechs2.txt','wb')
# for a in allSpeech:
# try:
# newfile2.write(a+' '+str(len(allSpeechTuple[a]))+'\n')
# except:
# print a
# newfile2.close()
print len(allSpeech)
def test(filename):
global count
fopen = open(filename)
content = fopen.read()
words = content.split('_')
for word in words:
mec = mecab.MeCabClass(word)
allw = mec.GetNoun2()
if len(allw) > 3:
count = count + 1
def chooseUse(filename, recordlist, recordfile):
global useCount
global uselessCount
item = Item()
record = Record()
fopen = open(filename, 'r')
content = fopen.read()[:-9]
paras = content.split("<b>○")
fopen.close()
for n in range(len(paras) - 1):
para_split = paras[n + 1].split("</b>")
#para_split[0]:speaker
#para_split[1]:speaking
item.speaker = para_split[0]
item.content = re.sub('<br>', '', para_split[1])
exit = ''
if len(item.content) < 300:
thisMecab = mecab.MeCabClass(item.content)
print item.content + " " + str(len(item.content))
print "Useful:", str(useCount), "Useless:", str(
uselessCount), "all is", str(useCount + uselessCount), '\n'
input = raw_input("y/n/e:")
while (input != "y" or input != "n"):
if input == "y":
record.length = len(item.content)
record.bool = 1
recordlist.append([record.length, record.bool])
recordfile.write(str(record.bool))
recordfile.write(":")
recordfile.write(item.content)
recordfile.write(">\n")
useCount += 1
break
elif input == "n":
record.length = len(item.content)
record.bool = 0
recordfile.write(str(record.bool))
recordfile.write(":")
recordfile.write(item.content)
recordfile.write(">\n")
uselessCount = uselessCount + 1
break
elif input == 'e':
print "you really want to exit it?:(y/n)"
exit = raw_input("y/n:")
if exit == "y":
return 1
else:
print "Please input again:(y/n/e):"
input = raw_input("y/n:")
print "\n\n\n"
def nounExtractFrom2000(fromFile,toFile):
# aimFilePath = "/home/dream/documents/papers/code/datatest/to2000Noun/"
fopen = open(fromFile,'r')
newfile = open(toFile,'wb')
text = fopen.read()
textMecab = mecab.MeCabClass(text)
allWordLists = textMecab.GetAll2()
for n in range(len(allWordLists)):
if textMecab.isNoun2(allWordLists[n][1]):
newfile.write(allWordLists[n][0])
if n != len(allWordLists) - 1:
if textMecab.isNoun2(allWordLists[n+1][1]) == 0:
newfile.write("_")
fopen.close()
newfile.close()
def speechStats2():
global TwoCooccurFile
global ThreeCooccurFile
# lines = Tools.content_lines(TwoCooccurFile)
wordListFre = Tools.content_scoreList(TwoCooccurFile, [':', ' '])
wordList = Tools.content_wordList(TwoCooccurFile, ':')
allSpeech = set()
allSpeechTuple = {}
allSpeechTuple2 = {}
for n in range(len(wordList) - 1):
# print wordList[n],wordListFre[n]
thisWordMec = mecab.MeCabClass(wordList[n])
thisSpeechs = thisWordMec.GetAll3()
try:
thisSpeech = thisSpeechs[0]
except:
pass
for x in range(len(thisSpeechs) - 1):
thisSpeech += "+" + thisSpeechs[x + 1]
try:
if thisSpeech in allSpeechTuple.keys():
allSpeechTuple2[thisSpeech] += wordListFre[n]
allSpeechTuple[thisSpeech] += 1
else:
allSpeechTuple2[thisSpeech] = wordListFre[n]
allSpeechTuple[thisSpeech] = 1
except:
print n, wordList[n]
sortedList = Tools.TupleSort(allSpeechTuple2)
Speech = []
SpeechCnt = []
Cnt = 0
for ele in sortedList:
Cnt += 1
# print Cnt,':',ele[0],' ',ele[1]
# print ele[0],'\t',ele[1]
Speech.append(Cnt)
SpeechCnt.append(ele[1])
title = "Part_of_speech-frequence distribution"
Tools.plot(Speech,
SpeechCnt,
title,
xlabel="Part of speech",
ylabel="frequence")
def getWordBox(self):
runCounter = 0
for file in self.fileLists:
runCounter += 1
if runCounter % 200 == 0:
os.system('clear')
print "This is the", runCounter, "th lines"
fopen = open(file, 'r')
content = fopen.read()
paras = content.split('>>>')
for para in paras:
nounInCon = mecab.MeCabClass(para.split('::')[-1]).GetNoun2()
self.counter += len(nounInCon)
for noun in nounInCon:
if noun not in self.wordBox:
self.wordBox[noun] = 1
else:
self.wordBox[noun] += 1
return self.wordBox
def nounExtractFromPretreat(filename,newfilename):
fopen = open(filename,'r')
text = fopen.read()
contents = text.split('>>>')
newfile = open(newfilename,'wb')
for content in contents:
speaking = content.split('::')[-1]
paras = speaking.split('\n')
for para in paras:
textMecab = mecab.MeCabClass(para)
allWordLists = textMecab.GetAll2()
for n in range(len(allWordLists)):
if textMecab.isNoun2(allWordLists[n][1]):
newfile.write(allWordLists[n][0])
if n != len(allWordLists) - 1:
if textMecab.isNoun2(allWordLists[n+1][1]) == 0:
newfile.write("_")
newfile.write("\n")
newfile.close()
fopen.close()
def isSpamMail(mail,spamDict,ordDict):
global spamPro
global ordPro
global spamNum
global ordNum
wordBox = {}
wordsNum = 0
wordInMail = mecab.MeCabClass(mail)
NounWords = wordInMail.GetNoun2()
AdjWords = wordInMail.GetAdj()
for noun in NounWords:
addWordToDict(noun,wordBox)
wordsNum = wordsNum + 1
for adj in AdjWords:
addWordToDict(adj,wordBox)
wordsNum = wordsNum + 1
spam = 1
ordinary = 1
#begin to get probility of per features
print len(wordBox)
for key in wordBox:
if key not in spamDict:
spam = spam * (0.5/spamNum)**(10/len(wordBox))
else:
spam = spam * (spamDict[key]/spamNum)**(10/len(wordBox))
if key not in ordDict:
ordinary = ordinary * (0.5/ordNum) **(10/len(wordBox))
else:
ordinary = ordinary * (ordDict[key] /ordNum) **(10/len(wordBox))
spam = spamPro*spam
ordinary = ordPro*ordinary
print "spam: ",spam
print "ordinary: ",ordinary
if spam > ordinary:
return True
else:
return False
def main1(tofile):
lines = Tools.content_lines("/home/dreamer/documents/code/eliminate/test/test_dict_for_singleWord.txt")
wordDictTuple = {}
for l in lines:
if l != '':
spl = l.split(',')
wordDictTuple[spl[1]] = spl[2][2:-2]
lines = Tools.content_lines("/home/dreamer/documents/code/eliminate/test/test.txt")
to = open(tofile,'wb')
for line in lines[:-1]:
try:
a = mecab.MeCabClass(line).GetAll()
to.write(a[0]+','+a[1]+',')
if a[0] not in wordDictTuple.keys() and a[1] not in wordDictTuple.keys():
to.write('\n')
elif a[0] not in wordDictTuple.keys():
to.write(','+str(wordDictTuple[a[1]])+'\n')
elif a[1] not in wordDictTuple.keys():
to.write(str(wordDictTuple[a[0]])+','+'\n')
else:
to.write(str(wordDictTuple[a[0]])+','+str(wordDictTuple[a[1]])+'\n')
except:
print line
to.close()
def generateMecabList(allWordFileName, topath):
allWordFile = open(allWordFileName, 'r')
content = allWordFile.read()
lines = content.split('\n')
OneFile = open(topath + 'OneWordsAdjacency.txt', 'wb')
TwoFile = open(topath + 'TwoWordsAdjacency.txt', 'wb')
ThreeFile = open(topath + 'ThreeWordsAdjacency.txt', 'wb')
MultiFile = open(topath + 'MultiWordsAdjacency.txt', 'wb')
AllFile = open(topath + 'AllWordsAdjacency.txt', 'wb')
runCounter = 0
for line in lines:
runCounter += 1
if runCounter % 2000 == 0:
os.system('clear')
print "This is the", runCounter, "th lines"
try:
lineEle = line.split('\t')
except:
continue
word = lineEle[0]
try:
wordFreStr = lineEle[1]
except:
continue
wordFre = int(wordFreStr)
if wordFre > 0:
mec = mecab.MeCabClass(word)
wordbox = mec.GetAll()
if len(wordbox) == 1:
OneFile.write(word)
AllFile.write(word)
OneFile.write(":")
AllFile.write(":")
OneFile.write(wordbox[0] + " " + wordFreStr + '\n')
AllFile.write(wordbox[0] + " " + wordFreStr + '\n')
elif len(wordbox) == 2:
TwoFile.write(word)
TwoFile.write(":")
TwoFile.write(wordbox[0] + ' ' + wordbox[1] + " " +
wordFreStr + '\n')
AllFile.write(word)
AllFile.write(":")
AllFile.write(wordbox[0] + ' ' + wordbox[1] + " " +
wordFreStr + '\n')
elif len(wordbox) == 3:
ThreeFile.write(word)
ThreeFile.write(":")
ThreeFile.write(wordbox[0] + ' ' + wordbox[1] + ' ' +
wordbox[2] + " " + wordFreStr + '\n')
AllFile.write(word)
AllFile.write(":")
AllFile.write(wordbox[0] + ' ' + wordbox[1] + ' ' +
wordbox[2] + " " + wordFreStr + '\n')
elif len(wordbox) > 3:
MultiFile.write(word)
MultiFile.write(":")
AllFile.write(word)
AllFile.write(":")
for index in range(len(wordbox) - 1):
MultiFile.write(wordbox[index] + ' ')
AllFile.write(wordbox[index] + ' ')
MultiFile.write(wordbox[-1] + " " + wordFreStr)
MultiFile.write('\n')
AllFile.write(wordbox[-1] + " " + wordFreStr)
AllFile.write('\n')
OneFile.close()
TwoFile.close()
ThreeFile.close()
MultiFile.close()
AllFile.close()
def divideIntoUseAndUnuse(TwoWordsMecabList, ThreeWordsMecabList, topath):
TwoWordsMecabListFile = open(TwoWordsMecabList, 'r')
ThreeWordsMecabListFile = open(ThreeWordsMecabList, 'r')
# MultiWordsMecabListFile = open(MultiWordsMecabList, 'r')
TwousefulSpeechList = Tools.content_wordList(
"/home/dreamer/documents/code/database/condition/SpeechUseful2.txt",
'\t')
ThreeusefulSpeechList = Tools.content_wordList(
"/home/dreamer/documents/code/database/condition/SpeechUseful3.txt",
' \t')
TwoUsefulFileName = topath + "TwoWordsAdjacency_afterSpeech.txt"
ThreeUsefulFile = topath + "ThreeWordsAdjacency_afterSpeech.txt"
TwoUsefulFile = open(TwoUsefulFileName, 'wb')
ThreeUsefulFile = open(ThreeUsefulFile, 'wb')
TwoWordsContent = TwoWordsMecabListFile.read()
TwoWordSet = set()
oneWordList = TwoWordsContent.split('\n')
allSpeechTuple2 = {}
for oneWord in oneWordList:
if oneWord == '':
continue
thisWord = oneWord.split(':')[0]
TwoWordSet.add(thisWord)
thisWordMec = mecab.MeCabClass(thisWord)
thisSpeechs = thisWordMec.GetAll3()
try:
thisSpeech = thisSpeechs[0]
for n in range(len(thisSpeechs) - 1):
thisSpeech += "+" + thisSpeechs[n + 1]
if thisSpeech in TwousefulSpeechList:
TwoUsefulFile.write(oneWord + '\n')
except:
pass
ThreeWordsContent = ThreeWordsMecabListFile.read()
WordList3 = ThreeWordsContent.split('\n')
allSpeechTuple3 = {}
for oneWord in WordList3:
if oneWord == '':
continue
thisWord = oneWord.split(':')[0]
thisWordMec = mecab.MeCabClass(thisWord)
thisSpeechs = thisWordMec.GetAll3()
try:
thisSpeech = thisSpeechs[0]
for n in range(len(thisSpeechs) - 1):
thisSpeech += "+" + thisSpeechs[n + 1]
if thisSpeech in ThreeusefulSpeechList:
ThreeUsefulFile.write(oneWord + '\n')
else:
pass
except:
traceback.print_exc()
continue
TwoWordsMecabListFile.close()
ThreeWordsMecabListFile.close()
TwoUsefulFile.close()
ThreeUsefulFile.close()
#!/usr/bin/python # -*- coding: utf-8 -*- import mecab text = "お隣中国" # mecab.MeCabClassForText(text) test = mecab.MeCabClass(text) # a = test.GetAll3() # print b # for b in a: print test.PrintAll() # 1. print # print test.PrintNoun2AndMark() # print test.PrintNoun2() # 2. get (retur n list) # VerbNodeList = test.GetAdj() # mydict = dict() # for node in VerbNodeList: # if node not in mydict: # mydict[node] = 1 # else: # mydict[node] += 1
def GetConnPro():
global TwoCooccurFile
global ThreeCooccurFile
TwoWordList = Tools.content_wordList(TwoCooccurFile, ':')
TwoWordCntlist = Tools.content_scoreList(TwoCooccurFile, [':', ' '])
# print len(AllWordList),len(AllWordCntlist),len(TwoWordList),len(TwoWordCntlist)
# print AllWordList[-1],AllWordCntlist[-1],TwoWordList[-1],TwoWordCntlist[-1]
TwoWordCnt = 0
for Cnt in TwoWordCntlist:
TwoWordCnt += Cnt
OneWordSpeechTuple = {}
TwoWordSpeechTuple = {}
# print key
for n in range(len(TwoWordList)):
thisWordMec = mecab.MeCabClass(TwoWordList[n])
thisSpeechs = thisWordMec.GetAll3()
thisSpeech = thisSpeechs[0] + "+" + thisSpeechs[1]
if thisSpeechs[0] in OneWordSpeechTuple:
OneWordSpeechTuple[thisSpeechs[0]] += TwoWordCntlist[n]
else:
OneWordSpeechTuple[thisSpeechs[0]] = TwoWordCntlist[n]
if thisSpeechs[1] in OneWordSpeechTuple:
OneWordSpeechTuple[thisSpeechs[1]] += TwoWordCntlist[n]
else:
OneWordSpeechTuple[thisSpeechs[1]] = TwoWordCntlist[n]
if thisSpeech in TwoWordSpeechTuple:
TwoWordSpeechTuple[thisSpeech] += TwoWordCntlist[n]
else:
TwoWordSpeechTuple[thisSpeech] = TwoWordCntlist[n]
# for key in TwoWordSpeechTuple.keys():
# # TwoWordSpeechTuple[key] = float(TwoWordSpeechTuple[key])/float(TwoWordCnt)
# print key,TwoWordSpeechTuple[key]
# for key in OneWordSpeechTuple.keys():
# # OneWordSpeechTuple[key] = float(OneWordSpeechTuple[key])/float(TwoWordCnt*2)
# print key,OneWordSpeechTuple[key]
TwoWordConnTuple = {}
for key in TwoWordSpeechTuple.keys():
try:
a = key.split('+')[0]
b = key.split('+')[1]
except:
print key
try:
c1 = OneWordSpeechTuple[a]
c2 = OneWordSpeechTuple[b]
except:
return
c12 = TwoWordSpeechTuple[key]
p = float(c2) / float(TwoWordCnt * 2)
p1 = float(c12) / float(c1)
p2 = float((c2 - c12)) / float((TwoWordCnt * 2 - c1))
try:
score = math.log(binomial(c12,c1,p)) + math.log(binomial(c2-c12,TwoWordCnt*2-c1,p)) \
- math.log(binomial(c12,c1,p1)) - math.log(binomial(c2-c12,TwoWordCnt*2-c1,p2))
# TwoWordConnTuple[key].write(wholeWord+'\t:'+subword1+' '+str(c1)+'\t:'+subword2+' '+str(c2)+'\t:'+str(c12)+'\t:'+str(score)+'\n')
TwoWordConnTuple[key] = score
except:
# traceback.print_exc()
# TwoScoreRecord.write(wholeWord+'\t:'+subword1+' '+str(c1)+'\t:'+subword2+' '+str(c2)+'\t:'+str(c12)+'\t:'+str(-9999)+'\n')
TwoWordConnTuple[key] = -9999
TwoWordConnList = Tools.TupleSort(TwoWordConnTuple)
Speech = []
SpeechCnt = []
Cnt = 0
for ele in TwoWordConnList:
Cnt += 1
print Cnt, ':', ele[0], ele[1], TwoWordSpeechTuple[ele[0]]
Speech.append(Cnt)
SpeechCnt.append(ele[1])
Tools.plot(Speech, SpeechCnt)