An Analysis Method of User Group's Emotional Tendency in Hot Topic of Microblog

Aoqiang Zhu, Shunxiang Zhang

Abstract: Microblog hot topic is the focus of microblog users’ attention. It is the key to master the emotional orientation of the public to accurately understand the emotional tendency of users to hot topics. How to accurately analyze the user group’s emotional tendentiousness is an urgent problem to be solved. This paper proposes an analysis method of user group’s emotional tendency of microblog hot topic based on the construction of exclusive dictionaries, which combines the construction of emotional dictionaries with hot topic. Firstly, use TF-IDF and K-means++ clustering to select the emotional words and characteristic words under the microblog, then, calculate the emotional intensity of the emotional words and build the exclusive Dictionary of this topic. Finally, we use the exclusive dictionary, based on the sentence level emotion analysis method, to analyze the emotional tendency of user groups. Experiments show that this method can significantly improve the accuracy of sentiment analysis of microblog hot topic users.