def coverage_values(self, sens_vector, essay_vector):
coverage_list = []
for sv in sens_vector:
coverage_list.append(self.dist.sim(sv, essay_vector,
self.disttype))
tools.normarlize(coverage_list)
return coverage_list
def clueswords_values(self, sens_words):
clue_list = [0] * len(sens_words)
words = []
for var in sens_words:
for w in var:
words.append(w)
sen_len_list = []
for i in range(len(sens_words)):
sen_w = sens_words[i]
sen_len_list.append(len(sen_w) / len(words))
for w in sen_w:
if w in self.cluewords:
clue_list[i] = 1
break
tools.normarlize(sen_len_list)
for i in range(len(sens_words)):
clue_list[i] = self.clue_weight * clue_list[i] + (
1 - self.clue_weight) * sen_len_list[i]
return clue_list
if __name__ == "__main__":
path = Dir.res + "/cleandata_small/news/trainning_2788.txt"
text = ftools.read_lines(path)
text = '。'.join(text)
asv = Auto_Simple_Vec()
sens, sens_words, sens_tags = asv.preprocess(text)
# for var in sens_words:
# print(var)
print("se_words lgth", len(sens_words))
sen_vec, essay_vec = asv.vectorize(sens_words, sens_tags)
# print(essay_vec)
print(sens[0], sens[1])
print(asv.dist.sim(sen_vec[0], sen_vec[1]))
print(asv.dist.sim(sen_vec[0], sen_vec[-1]))
print(asv.dist.sim(sen_vec[2], sen_vec[3]))
coverage_list = []
for i in range(len(sen_vec)):
# print(sen_vec[i])
# input()
coverage_list.append(asv.dist.sim(sen_vec[i], essay_vec, Distance.EUD))
tools.normarlize(coverage_list)
for i in range(len(coverage_list)):
print(sens[i], ",", coverage_list[i])
# print("------en--------")
# print(essay_vec)