Prediction of Emotional Tendency of COVID-19 Speech Based on Pre-training Model

Abstract

In 2020, the World Health Organization confirmed the new coronavirus (COVID-19) as a pandemic infectious disease. This virus affects the lives of tens of millions of people. Many countries have adopted compulsory measures to require residents to work and live at home and avoid unnecessary contact with the outside world. During this period, people expressed their feelings about the epidemic through various online platforms. In this article, we conducted an effective sentiment analysis on Weibo. These Weibo data are from January 1, 2020 to February 20, 2020, all related to COVID-19. The sentiment analysis of epidemic speech is to analyze people's views on COVID-19 during the epidemic. The results of the study concluded that although the epidemic is gradually getting worse, most people don't take a negative attitude.