def getDelRefMap():
delRefList = getdelRef()
chosedRef = getRefList()
WordEmbedding = getEmbedding.getEmbeddingMat_senna()
SennaList = getSennaList()
refmap = []
# count = 0
for i in range(100000, 120000):
word1 = delRefList[i]
print(i)
# count += 1
if i % 100 == 0:
wfile = open("res/sennaMap_4739_E.txt", "w")
wfile.writelines(item + '\n' for item in refmap)
if word1 not in SennaList:
continue
w1vec = WordEmbedding[SennaList.index(word1), :]
sim = []
for word2 in chosedRef:
if word2 not in SennaList:
sim.append(0)
continue
w2vec = WordEmbedding[SennaList.index(word2), :]
sim.append(vecSim(w1vec, w2vec))
chosed = chosedRef[np.argmax(sim)]
item = word1 + "\t" + chosed
print(item)
refmap.append(item)
wfile = open("res/sennaMap_4739_E.txt", "w")
wfile.writelines(item + '\n' for item in refmap)
def getDelRefMap():
delRefList = getdelRef()
chosedRef = getRefList()
WordEmbedding = getEmbedding.getEmbeddingMat_senna()
SennaList = getSennaList()
refmap = open("res/sennaMap_4739_1-20000.txt").readlines()
# count = 0
# start = len(refmap)
for i in range(50000,100000):
word1 = delRefList[i]
print(i)
# count += 1
if i%100 == 0:
wfile = open("res/sennaMap_4739_D.txt","w")
wfile.writelines(item+'\n' for item in refmap)
if word1 not in SennaList:
continue
w1vec = WordEmbedding[SennaList.index(word1),:]
sim = []
for word2 in chosedRef:
if word2 not in SennaList:
sim.append(0)
continue
w2vec = WordEmbedding[SennaList.index(word2),:]
sim.append(vecSim(w1vec,w2vec))
chosed = chosedRef[np.argmax(sim)]
item = word1+"\t"+chosed
print(item)
refmap.append(item)
wfile = open("res/sennaMap_4739_D.txt","w")
wfile.writelines(item+'\n' for item in refmap)
def getWordPairVec(w1,w2):
# root = u"I:/数据/word12585relation30/rel_30_ref_TFIDF/ref_800_TFIDF/rel_spc/spc_1_350/word_matrix/"
wordList = getWordList()
WordEmbedding = getEmbedding.getEmbeddingMat_senna()
w1vec = WordEmbedding[wordList.index(w1),:]
w2vec = WordEmbedding[wordList.index(w2),:]
vec = hstack([w1vec,w2vec])
# print np.shape(w1Mat)[1]
return vec
def getWordPairVec(w1, w2):
# root = u"I:/数据/word12585relation30/rel_30_ref_TFIDF/ref_800_TFIDF/rel_spc/spc_1_350/word_matrix/"
wordList = getWordList()
WordEmbedding = getEmbedding.getEmbeddingMat_senna()
w1vec = WordEmbedding[wordList.index(w1), :]
w2vec = WordEmbedding[wordList.index(w2), :]
vec = hstack([w1vec, w2vec])
# print np.shape(w1Mat)[1]
return vec
def getDelRefMap():
delRefList = getdelRef()
chosedRef = getRefList()
WordEmbedding = getEmbedding.getEmbeddingMat_senna()
SennaList = getSennaList()
map = []
for word1 in delRefList:
if word1 not in SennaList:
continue
w1vec = WordEmbedding[SennaList.index(word1), :]
sim = []
for word2 in chosedRef:
if word2 not in SennaList:
sim.append(0)
continue
w2vec = WordEmbedding[SennaList.index(word2), :]
sim.append(vecSim(w1vec, w2vec))
chosed = chosedRef[np.argmax(sim)]
print(word1 + "\t" + chosed)
map.append(word1 + "\t" + chosed)
wfile = open("res/sennaMap6830.txt", "w")
wfile.writelines(item + '\n' for item in map)
def getDelRefMap():
delRefList = getdelRef()
chosedRef = getRefList()
WordEmbedding = getEmbedding.getEmbeddingMat_senna()
SennaList = getSennaList()
map = []
for word1 in delRefList:
if word1 not in SennaList:
continue
w1vec = WordEmbedding[SennaList.index(word1),:]
sim = []
for word2 in chosedRef:
if word2 not in SennaList:
sim.append(0)
continue
w2vec = WordEmbedding[SennaList.index(word2),:]
sim.append(vecSim(w1vec,w2vec))
chosed = chosedRef[np.argmax(sim)]
print(word1+"\t"+chosed)
map.append(word1+"\t"+chosed)
wfile = open("res/sennaMap6830.txt","w")
wfile.writelines(item+'\n' for item in map)
def vecSim(x, y):
result1 = np.dot(x,y)
result2 = np.linalg.norm(x)
result3 = np.linalg.norm(y)
if(result2*result3 == 0):
return 1
cos = result1/(result2*result3)
sim = 0.5+0.5*cos
return sim
delRefList = getdelRef()
chosedRef = getRefList()
WordEmbedding = getEmbedding.getEmbeddingMat_senna()
SennaList = getSennaList()
refmap = []
# count = 0
def getDelRefMap(lock,index):
count = 0
for i in range(index,index*30000):
word1 = delRefList[i]
# for word1 in delRefList:
# print(count)
# count += 1
print(count)
if word1 not in SennaList:
rfile.close()
wfile = open(u"res/GS2011_SENNA.txt", "w")
wfile.writelines(item for item in sampleList)
def getWordList():
rfile = open(u"I:/数据/embedings/senna/words.txt")
wordList = []
for line in rfile:
wordList.append(line.strip("\n"))
rfile.close()
# arr = np.array(wordList)
return wordList
WordEmbedding = getEmbedding.getEmbeddingMat_senna()
wordList = getWordList()
def sentCombMat_Mult(w1, w2, w3):
w1Mat = WordEmbedding[wordList.index(w1), :]
w2Mat = WordEmbedding[wordList.index(w1), :]
w3Mat = WordEmbedding[wordList.index(w1), :]
return w1Mat * w2Mat * w3Mat
def vecSim(x, y):
result1 = np.dot(x, y)