def calculate(givenWord,loop,first,second):
wordId1 = getId.main(stemming.main(givenWord))
for tweetId in tweetIdList[first:second]:
oneTweetWords = tweetId_words[tweetId]
for word in oneTweetWords:
wordId2 = getId.main(stemming.main(word))
co = cooccurrences.main(wordId1,wordId2)
if co > 0:
similarityWord = getWordsSimilarity.main(wordId1,wordId2)
if similarityWord == None:
vector2 = json.load(open("/data/CikmTwitterProject/Paper/SearchTweets/WordsDistribution#5/"+str(wordId2)))
similarityWord = countSim(vector1,vector2)
storeWordsSimilarity.main(wordId1,wordId2,similarityWord)
def main(word):
print "Start:",word
wordId = getId.main(word)
if wordId != None:
distributionList = getDistribution.main(wordId)
else:
distributionList = []
vector = {}
for i in range(len(distributionList)):
vector[i+1] = distributionList[i]
words = vector.keys()
values = vector.values()
values.sort(reverse=True)
fd2 = open("topN/" + word +"_Top" + str(n),"a+")
if len(values) > 0:
for i in range(n):
if float(values[i]) != float(0):
id = vector.keys()[vector.values().index(values[i])]
vector.pop(id)
word = str(id)
fd2.write(word)
fd2.write(",")
fd2.write(str(values[i]))
fd2.write("\n")
else:
break
def main(givenWord):
db_mysql = MySQLdb.connect('141.117.3.92','lunafeng','luna222','CikmTwitterDataSet')
wordId = getId.main(stemming.main(givenWord))
global tweetIdList
global tweetId_words
global vector1
tweetId_words = {}
while(db_mysql.open != 1):
db_mysql.ping()
cursor = db_mysql.cursor()
# removed words itself and love, like, day, lol, today, tomorrow, time, tonight, thing, ready, found,free
getTweetWords = "select A.tweetId,w.word from Words w join (select c.tweetId,tw.wordId from " + givenWord + " c join TweetsWords tw on (c.TweetId = tw.TweetId)) A on (w.id = A.wordId) where A.wordId !=" + str(wordId) + " and A.wordId != 1804507 and A.wordId != 1040690 and A.wordId != 1111170 and A.wordId != 991563 and A.wordId != 13304 and A.wordId != 3368935 and A.wordId != 2113819 and A.wordId != 1990840 and A.wordId != 2977454 and A.wordId != 3489500 and A.wordId != 1326944 and A.wordId != 419686"
cursor.execute(getTweetWords)
resultsRaw = cursor.fetchall()
cursor.close()
db_mysql.close()
for result in resultsRaw:
tweetId = result[0]
word = result[1]
if tweetId not in tweetId_words:
tweetId_words[tweetId] = []
tweetId_words[tweetId].append(word)
vector1 = json.load(open("/data/CikmTwitterProject/Paper/SearchTweets/WordsDistribution#5/"+str(wordId)))
pids = []
tweetIdList = tweetId_words.keys()
print "length:",len(tweetIdList)
time.sleep(1)
tweetIdList.sort()
calculate(givenWord,0,0,len(tweetIdList))
def main(word):
filePath = "distributionList/"
db_mysql = MySQLdb.connect('141.117.3.92','lunafeng','luna222','WordsDisambiguation_b4')
Id1 = getId.main(word)
if not os.path.exists("distributionList/" + str(Id1) + ".List"):
db_mysql.ping(True)
cursor = db_mysql.cursor()
getAllWords = "select Word from TotalOccurrencesAll"
cursor.execute(getAllWords)
allWords = cursor.fetchall()
cursor.close()
global element_list
element_list = []
for element in allWords:
element_list.append(int(element[0]))
matrix_hash = json.load(open("../Import/MatrixCPNoOneNoCommonNoSame_b4"))
walkNumber = 5
vector1 = stationary_dist(Id1,matrix_hash,element_list,walkNumber)
fd = open(filePath + str(Id1) + ".List","w+")
valueList = []
for i in range(28498):
try:
value = vector1[int(i+1)]
except:
value = ''
fd.write(str(value) + "\n")
fd.flush()
db_mysql.close()
def main(word):
filePath = "distributionList/"
db_mysql = MySQLdb.connect('141.117.3.92', 'lunafeng', 'luna222',
'WordsDisambiguation_b4')
Id1 = getId.main(word)
if not os.path.exists("distributionList/" + str(Id1) + ".List"):
db_mysql.ping(True)
cursor = db_mysql.cursor()
getAllWords = "select Word from TotalOccurrencesAll"
cursor.execute(getAllWords)
allWords = cursor.fetchall()
cursor.close()
global element_list
element_list = []
for element in allWords:
element_list.append(int(element[0]))
matrix_hash = json.load(
open("../Import/MatrixCPNoOneNoCommonNoSame_b4"))
walkNumber = 5
vector1 = stationary_dist(Id1, matrix_hash, element_list, walkNumber)
fd = open(filePath + str(Id1) + ".List", "w+")
valueList = []
for i in range(28498):
try:
value = vector1[int(i + 1)]
except:
value = ''
fd.write(str(value) + "\n")
fd.flush()
db_mysql.close()
def calculate(givenWord,loop,first,second):
fd = open("Luna"+givenWord+"RemoveCo0"+str(loop),"w+")
wordId1 = getId.main(stemming.main(givenWord))
for tweetId in tweetIdList[first:second]:
similarityFinal = 0
oneTweetWords = tweetId_words[tweetId]
length = 0
length = len(oneTweetWords)
for word in oneTweetWords:
wordId2 = getId.main(stemming.main(word))
co = cooccurrences.main(wordId1,wordId2)
if co > 0:
similarityWord = getWordsRefinedSimilarity.main(wordId1,wordId2)
similarityFinal += float(similarityWord)
similarityFinal = float(similarityFinal)/length
fd.write(str(tweetId))
fd.write("\t")
fd.write(str(float(similarityFinal)))
fd.write("\n")
fd.flush()
def main(givenWord):
wordId = getId.main(givenWord)
wordsAll = []
if os.path.exists("filtered/" + givenWord + "_Top" + str(n)):
top = open("filtered/" + givenWord + "_Top" + str(n), "r")
rawList = top.readlines()
for raw in rawList:
wordId = raw.strip("\n")
if not os.path.exists("distributions/" + str(wordId) + ".List"):
wordsAll.append(wordId)
return wordsAll
def main(givenWord):
wordId = getId.main(stemming.main(givenWord))
db_mysql = MySQLdb.connect('141.117.3.92','lunafeng','luna222','CikmTwitterDataSet')
db_mysql.ping()
cursor = db_mysql.cursor()
Max = 2.62507
update = "UPDATE WordsSimilarity SET RefinedSimilarity=(" + str(Max) + "-ABS(Log(10,Similarity)))*10 WHERE RefinedSimilarity IS NULL"
print update
cursor.execute(update)
db_mysql.commit()
cursor.close()
db_mysql.close()
def main(givenWord):
wordId = getId.main(stemming.main(givenWord))
db_mysql = MySQLdb.connect('141.117.3.92', 'lunafeng', 'luna222',
'CikmTwitterDataSet')
db_mysql.ping()
cursor = db_mysql.cursor()
Max = 2.62507
update = "UPDATE WordsSimilarity SET RefinedSimilarity=(" + str(
Max) + "-ABS(Log(10,Similarity)))*10 WHERE RefinedSimilarity IS NULL"
print update
cursor.execute(update)
db_mysql.commit()
cursor.close()
db_mysql.close()
def main(givenWord):
wordId = getId.main(stemming.main(givenWord))
global tweetIdList
global tweetId_words
global vector1
tweetId_words = {}
db_mysql = MySQLdb.connect('141.117.3.92','lunafeng','luna222','CikmTwitterDataSet')
db_mysql.ping()
cursor = db_mysql.cursor()
getTweetWords = "select A.tweetId,w.word from Words w join (select c.tweetId,tw.wordId from " + givenWord + " c join TweetsWords tw on (c.TweetId = tw.TweetId)) A on (w.id = A.wordId) where A.wordId !=" + str(wordId)
cursor.execute(getTweetWords)
resultsRaw = cursor.fetchall()
for result in resultsRaw:
tweetId = result[0]
word = result[1]
if tweetId not in tweetId_words:
tweetId_words[tweetId] = []
tweetId_words[tweetId].append(word)
cursor.close()
db_mysql.close()
vector1 = json.load(open("/data/CikmTwitterProject/Paper/SearchTweets/WordsDistribution#5/"+str(wordId)))
pids = []
tweetIdList = tweetId_words.keys()
tweetIdList.sort()
for loop in range(8):
pid = os.fork()
pids.append(pid)
if pid == 0 :
first = loop*len(tweetIdList)/8
second = (loop+1)*len(tweetIdList)/8
calculate(givenWord,loop,first,second)
os._exit(0)
else:
continue
for pid in pids:
os.waitpid(pid,0)
#!/usr/bin/python
import MySQLdb
import getId, stemming
def main(word1, word2):
db_mysql = MySQLdb.connect('141.117.3.92', 'lunafeng', 'luna222',
'CikmTwitterDataSet')
db_mysql.ping()
cursor = db_mysql.cursor()
sql = "SELECT Similarity FROM WordsSimilarity WHERE Word1=" + str(
word1) + " AND Word2=" + str(word2)
try:
cursor.execute(sql)
similarity = cursor.fetchone()
except:
db_mysql.rollback()
cursor.close()
db_mysql.close()
try:
similarity = similarity[0]
return similarity
except:
return None
words1 = getId.main(stemming.main("popcorn"))
words2 = getId.main(stemming.main("microwave"))
print main(words1, words2)
#!/usr/bin/python
import MySQLdb
import getId,stemming
def main(word1,word2):
db_mysql = MySQLdb.connect('141.117.3.92','lunafeng','luna222','CikmTwitterDataSet')
db_mysql.ping()
cursor = db_mysql.cursor()
sql = "SELECT Similarity FROM WordsSimilarity WHERE Word1=" + str(word1) + " AND Word2=" + str(word2)
try:
cursor.execute(sql)
similarity = cursor.fetchone()
except:
db_mysql.rollback()
cursor.close()
db_mysql.close()
try:
similarity = similarity[0]
return similarity
except:
return None
words1 = getId.main(stemming.main("popcorn"))
words2 = getId.main(stemming.main("microwave"))
print main(words1,words2)
