Opinion mining and sentiment analysis model of microblog

摘要

In order to fulfill the task of recognition and sentiment analysis of perspective sentences from microblog, we propose an algorithm which based on Decision Tree composed by Neural Network and Bayes. Neural Network Algorithm was specialized for find sentences which contain viewpoint of users. In this part, we selected some group of features on Chinese articles. We also used corpus of Tan Songbo for data training. Bayes was specialized for sentiment analysis and trained by using corpus of Tan Songbo. The experiments we did at the end of this article demonstrates that, the algorithm perform excellent in accuracy rate.