Sentiment Analysis of Chinese Tourism Review Based on Boosting and LSTM
Pengfei Liu, Duxian Nie, Xiaxu He, Weifeng Zhang, Zhirui Huang, Kejing He · 2019
Sentiment analysis is a hot topic in recent years' Natural Language Processing Research, traditional practices are through machine learning, current research is based on deep learning. Most models generally do not consider the characteristic of network data while in modeling procedure. Considering the tourism comments emotion distribution is not balanced, and adding distribution information of comments to the model will play an important role in the effect, this paper proposed a boosting based LSTM model that can well model the comment text. We have done the evaluation on the related data set, and the experimental results show that the model in this paper has achieved good results.