In order to study the influence of university closure and prevent negative public opinion events, this paper crawls blog posts on the topic of "school closure" based on Weibo and analyzes their influence by fusing the sentiment of the text. The EMBUI algorithm was constructed to calculate the fusion influence. In EMBUI, participant influence is calculated based on the MBUInfluence algorithm and sentiment feature is incorporated, which are obtained by machine learning methods. The experimental results show that the fusion influence can effectively grasp the public opinion, and Logistic Regression works best in sentiment analysis. The influence of the participants is related to their own characteristics and the influence of their fans, and there is a strong correlation between their influence and the sentiment. Influence has different sentiment features, which need to be captured separately.