Trends, Politics, Sentiments, and Misinformation: Understanding People's Reactions to COVID-19 During its Early Stages

The sudden outbreak of COVID-19 resulted in large volumes of data shared on different social media platforms. Analyzing and visualizing these data is doubtlessly essential to having a deep understanding of the pandemic's impacts on people's lives and their reactions to them. In this work, we conduct a large-scale spatiotemporal data analytic study to understand peoples' reactions to the COVID-19 pandemic during its early stages. In particular, we analyze a JSON-based dataset that is collected from news/messages/boards/blogs in English about COVID-19 over a period of 4 months, for a total of 5.2M posts. The data are collected from December 2019 to March 2020 from several social media platforms such as Facebook, LinkedIn, Pinterest, StumbleUpon and VK. Our study aims mainly to understand which implications of COVID-19 have interested social media users the most and how did they vary over time, the spatiotemporal distribution of misinformation, and the public opinion toward public figures during the pandemic. Our results can be used by many parties (e.g., governments, psychologists, etc.) to make more informative decisions, taking into account the actual interests and opinions of the people.