Enhanced Public Opinion Analysis Method for Micro Blog Topic

Zhen-fei WANG,Kai-li LIU,Zhi-yun ZHENG,Yu SONG
2017-01-01
Abstract:As a new social networking,microblog is changing our lifestyle,involved increasingly the areas of community,and now many polls is completed by the public opinion analysis of microblog topic.With development of social network,such as microblog,people release their feelings and opinions on the micro-blog and other social platforms,Opinion polls now can be done through the analysis of the public opinion on micro-blog topics.The article in view of the existing research objects of sentiment analysis is relatively single,sentiment analysis,through the multidimensional association oriented data,fusion microblog topic diffusion degree and emotional tendencies for microblog topics of public opinion value enhance the validity of the results.Microblog topic sentiment analysis will take into account emoticons,build emoticons libraries through FP growth algorithm and searching distance,using Naive Bayesian method to complete sentiment analysis.On sina micro-blog network experiment data set,and the recall ratio and precision as the evaluation criteria.The results show that the method accuracy is up to 82.6%.
What problem does this paper attempt to address?