[Objective] The model is used to divide the user groups under the hot topics of Weibo.[Methods] Starting from the perspective of sentiment analysis, the sentiment dictionary method is used to calculate the value of users’s text sentiment. the text sentiment value has been combined with user text vector expression to construct the user's opinion sentiment characteristics. Then, the user's opinion sentiment characteristics have been used to divide user groups by the K-means method.[Results] The user group classification model proposed in this paper divides users into 3 categories, and the value of the evaluation index CA is 78.2%.[Limitations] The model needs to first determine the number of categories when dividing user groups.[Conclusions] The model constructed in this paper was effective. Users with the same sentimental views can be gathered together by classify model proposed in this paper.
张梦瑶, 朱广丽, 张顺香, 张标.
[J]. 数据分析与知识发现, 10.11925/infotech.2096-3467.2020.1059.
Zhang Mengyao, Zhu Guangli, Zhang Shunxiang, Zhang Biao.
A User Group Classification Model Based on Sentiment Analysis under Microblog Hot Topics
. Data Analysis and Knowledge Discovery, 0, (): 1-.