Let the Big Data Speak: Collaborative Model of Topic Extract and Sentiment Analysis COVID-19 Based on Weibo Data

Tianjie Luo, Ran Li, Zhe Sun, Fuqiang Tao, Manoj Kumar, Chao Li

Abstract: Micro-blog is an important medium of emergency communication. The topic and emotion analysis of micro-blog is of great significance in identifying and predicting potential problems and risks. In this paper, a collaborative analysis model of emotion and topic mining is constructed to analyze the users’ sentiment and the topics they care about, Firstly, we use SO-PMI to construct domain sentiment lexicon and extract topics with LDA. Then we use the collaborative model to analyze sentiment and topic. The results showed that the model we proposed can present the features of sentiment and topic of user concerns. And through text clustering and sentiment analysis, it is found that the attitude of users towards the COVID-19 has gone through three stages, namely, a period of fluctuating tension and anxiety, a period of slowly rising solidarity and a period of stable self-confidence with little fluctuation, on the whole, positive is greater than negative, positive than negative state.