Professor Choi Yong-seok (Department of Computer Science) and Son Dong-young (Department of Media Communication) from the Hanyang University's Center for Computational Social Science recently released the results of analysis on the relationship between the changes in the number of COVID-19 patients and people's anxiety.

The graph above shows 236,592 Twitter data posted between October 2018 and March 2021 based on a deep learning model and shows the average daily anxiety and the number of confirmed COVID-19 cases in time series. The blue line shows the number of confirmed COVID-19 cases since January 20th in 2020, when the first COVID-19 occurred, and the red line shows the average trend of anxiety. The black dotted line indicates the start of the 3 outbreaks in 2020 (February, August and November) in which the number of confirmed COVID-19 cases increased rapidly.

The average of anxiety, which soared nationwide during the 1st outbreak, gradually decreased, and showed a temporary rise in August when the 2nd outbreak occurred and November when the 3rd outbreak occurred. The level of emotional response to the sudden increase in the number of confirmed cases decreased day by day, and the anxiety level showed a long-term downward trend.

In particular, despite the rapid increase in the number of confirmed cases this year, the continuous decline in the anxiety index reflects the accumulated fatigue and decreased sensitivity of infectious diseases due to prolonged COVID-19. Such a change can be seen as an indication that communication and policy responses related to quarantine by the government may become more difficult.

Professor Choi Young-seok, who led the research, said, "The result is meaningful in that we were able to visually confirm the real-time changes in the level of anxiety nationwide about the COVID-19 situation."