A simple implement of deep learning in text classification.
This repository includes four models written in Keras: Bayes, SVM, CNN, ann LSTM. All of them could be run independently, and feel free to combine them. It works well!
I have pre-trainned a W2V model with huge Chinese corpus (dim=128, size=3.1G). Ask me if you need it. Or you could use any word embedding models instead.
I have crawled over 40,000 reviews from Taobao.com, the e-commerce leader in China. This project is built to detect review spam in that dataset. Also, ask me if you need it.
This repository is not creative but stable and useful. However, if you are interested in natural language processing and text classification, please feel free to talk to me. I have implemented a new project for review spam detection, in which I improve the BERT model, which is also generalizable to any text classification tasks. I will get it pulic later.