this is a simple TF-IDF algorithm that use python open source package "JIEBA" to cut Chinese character string into individual word, then use TfidfTransformer from sklearn to calcullate the TF-IDF value for each word in each setences.

Unofficial simplified implementation of the Contrastive unpaired translation paper, using TensorFlow and Keras.