(self, words, del_stopword=False)
| 43 | return self.w2v_model[word] |
| 44 | |
| 45 | def _predict(self, words, del_stopword=False): |
| 46 | word_list = jieba.cut(words,cut_all=False) |
| 47 | if del_stopword: |
| 48 | word_list = self.stop_word.del_stopwords(word_list) |
| 49 | zero_vec = np.zeros(256) |
| 50 | word_vector_list = [] |
| 51 | for word in word_list: |
| 52 | try: |
| 53 | word_vector_list.append( self.w2v_model[word] ) |
| 54 | except: |
| 55 | word_vector_list.append(zero_vec) |
| 56 | word_vector_list = np.array(word_vector_list).mean(axis=0) |
| 57 | return word_vector_list[np.newaxis, :].astype(float) |
| 58 | |
| 59 | |
| 60 | class Word2Vec(ModelBase): |
no test coverage detected