# 文本相似度評(píng)估方式
def exponent_neg_manhattan_distance(sent_left, sent_middle, sent_right):
  '''基于曼哈頓空間距離計(jì)算兩個(gè)字符串語(yǔ)義空間表示相似度計(jì)算'''
  return ((K.exp(-K.sum(K.abs(sent_left - sent_middle), axis=1, keepdims=True)) - K.exp(-K.sum(K.abs(sent_left - sent_right), axis=1, keepdims=True))) + 1) / 2
 
def bilstm_siamese_model():
  '''搭建孿生網(wǎng)絡(luò)'''
  #可以在這里調(diào)參
  embedding_layer = Embedding(VOCAB_SIZE + 1,
                EMBEDDING_DIM,
                weights=[embedding_matrix],
                input_length=MAX_LENGTH,
                trainable=True, #原本為False
                mask_zero=True)
  #輸入層
  left_input = Input(shape=(MAX_LENGTH,), dtype='float32', name="left_x") #(?, 25)
  middle_input = Input(shape=(MAX_LENGTH,), dtype='float32', name="middle_x") #(?, 25)
  right_input = Input(shape=(MAX_LENGTH,), dtype='float32', name='right_x')
  #嵌入層
  encoded_left = embedding_layer(left_input) #(?, 25, 300)
  encoded_middle = embedding_layer(middle_input) #(?, 25, 300)
  encoded_right = embedding_layer(right_input)
  # print(encoded_left)
  #孿生網(wǎng)絡(luò)
  shared_lstm = create_base_network(input_shape=(MAX_LENGTH, EMBEDDING_DIM))
  left_output = shared_lstm(encoded_left)
  middle_output = shared_lstm(encoded_middle)
  right_output = shared_lstm(encoded_right)
 
  # 文本相似度評(píng)估方式
  distance = Lambda(lambda x: exponent_neg_manhattan_distance(x[0], x[1], x[2]),output_shape=lambda x: (x[0][0], 1))([left_output, middle_output, right_output]) #第二個(gè)lambda函數(shù)的輸入?yún)?shù)不清楚，x[0][0]
  print('distance',distance)
  model = Model([left_input, middle_input, right_input], distance)
  model.compile(loss='binary_crossentropy',
         optimizer='nadam',
         metrics=['accuracy'])
  model.summary()
  return model