Chinese Micro-blog Sentiment Analysis System Based On Sentence Structure

With the rapid development of the Internet and the popularization of the mobile terminal, social network is more and more effects on human beings. Micro-blog is a kind of convenient social network who has very strong instantaneity. Gradually it comes into the lives of our lives. More and more people prefer to share information, access information and exchange feelings and ideas through the micro-blog. Because of its strong originality and being closer to the public. Emotion analysis on micro-blog which is a kind of short text can benefit a lot like monitoring public opinions.Sentiment analysis is that we do some further research based on the analysis of emotional tendency. In this system, we design seven sentiments, such as anger, disgust, fear, joy, sadness and surprise. Chinese micro-blog is different from English micro-blog and also different from written language. We design this system based on this characteristic. The data comes from sina micro-blog. This system get some micro-blogs including some basic information, such as when and where the micro-blog releases, using the API provided by the sina micro-blog. Then after some preprocessing on the micro-blog, we do word segmentation and analyzing sentence structure using the ICTACLAS system and ltp system. Finally, we use these result and the emotion and adverb word library to do sentiment analyzing. And store the result into MySQL database.Using jsp and tomcat, the result will be shown to the clients in line chart, bar chart and pie chart. Users can also check different results from different time, micro-blog from different places and different terminals. For the login users, they can check some further result.