Detecting Social Distress during COVID-19 Pandemic: Deep-Learning Classification and Tracking of Anxiety in Social-Media

초록

During a crisis like COVID-19 pandemic, individuals might respond psychologically to increased uncertainty. These responses are often manifested as expressions of anxiety, which reveal how individuals experience uncertainty in social environments. Identifying and tracking anxiety may enable us to understand how people in society collectively cope with social crises. This study aimed to develop deep-learning-based classifier to extract anxietyladen messages from Twitter and examine how the trend of anxiety corresponded to major waves of COVID-19.