Closed perveil closed 2 years ago
Hi @perveil ,
For document classification, we directly used the learned word embedding by CatE as the input word representation to WeSTClass (i.e., replacing the word2vec embedding in WeSTClass with the CatE embedding learned on the same corpus).
Best, Yu
hi yumeng: i am a new researcher in this field,i want to know after this model,how to do a classification task!in the paper,did you use the model of WeSTClass to classify the type of document?
BEST
a student from PKU.