Research on emotional analysis of netizens and topic distribution under public health emergencies : ——A Case Study of COVID-19

Abstract

[Purpose/meaning] The COVID-19 epidemic that has swept the world has caused people to fall into fear. It conducts sentiment analysis on netizens under public health emergencies and provides a reference for the government to sort out netizens' emotions during the epidemic. [Method/Procedure] LSTM sentiment classification model is built based on deep learning technology, sentiment analysis is carried out on the comments of netizens in Weibo, and the topic distribution of different sentiments of netizens is studied based on the LDA topic model. [Results/Conclusion] The results show that the negative emotions of netizens are about the same as positive emotions. Most positive emotions pay tribute to medical staff, and most of the negative emotions focus on the problem of not being able to buy a mask.