Multi-channel LSTM-CNN model for Vietnamese sentiment analysis
Quan-Hoang Vo, Huy Tien Nguyen, Bac Hoai Le, Le-Minh Nguyen · 2017
Convolutional neural network (CNN) and Long Short Term Memory (LSTM) have shown the state of the art results for sentiment analysis in English corpus. However, there are not many studies of this approach for Vietnamese corpus. In our work, CNN and LSTM are employed to generate information channels for Vietnamese sentiment analysis. Because each deep learning model (e.g. CNN, LSTM) has a particular advantage, this scenario provides a novel and efficient way for integrating the advantages of CNN and LSTM. In addition, we introduced a Vietnamese corpus, which collected comments/reviews from Vietnamese commercial web pages and was annotated by three human annotators. We evaluated our approach on our corpus and VLSP corpus. According to the experimental results, the proposed model outperforms SVM, LSTM, and CNN on the two datasets.