| In recent years, microblog has become one of the world’s most popular network applications, the rapid development of microblog has shown significant social value and commercial value, people gradually accustomed to obtain information in SNS such as microblog, also known as weibo. The text sentiment analysis focuses on the emotional tendencies of the text information and begin to play a role when a huge amount of data makes it impossible to analyze them manually. Sentiment analysis in the English-speaking world has been a wide range of research areas while the Chinese sentiment analysis research is still in its infancy. Most of the work by the attempt has been proved to be suitable for English.First, this paper summarizes and analyzes the basic concepts and algorithms model of text sentiment analysis. On the basis, this paper introduces the psychology of the PAD emotion model, combined with the semantic similarity calculation method provided by HowNet, proposed a method of using a given basic emotional vocabulary words and their corresponding PAD values, and build a dictionary based on it. Secondly, this paper extend the problem to the field of Chinese text sentiment analysis, proposes a combination of statistical information and semantic information weight calculation method, eliminate ambiguity characteristic for classification of such characteristics weights more fitting text semantic classification better. Finally, comprehending Chinese microblogging text analysis, this paper analyzes the existing text representation model that combines machine learning support vector machine algorithm is proposed based on semantic features PAD emotional support vector machine classification method.Experimental results show that results based on support vector machine classification algorithm is better than k-nearest neighbor algorithm. Meanwhile, this paper, based Semantic PAD emotional support vector machine classification method can obtain more practical results, and the general effect of the support vector machine approach has significant improvements. |
