ANALYSIS OF COVID-19 MISINFORMATION: ORIGIN AND CURE NARRATIVES

Overview

The interplay between three critical aspects --- biology of the virus, human behavior, and government policies --- significantly determines the spread and impact of any given epidemic. With over 5.5 million deaths worldwide, COVID-19 has exposed the vulnerabilities of modern societies to not only an epidemic outbreak but the corresponding spread of misinformation on social media platforms. In order to understand the human behavior that facilitated the rapid spread of COVID-19, we study the spread of misinformation through social media. While the biology governing the spread of a virus through a population is well understood, the spread of misinformation is relatively less understood, and therefore difficult to control. By selecting different narratives on the origin of COVID-19 and its unsubstantiated cures, we collected data from Reddit posts and their corresponding comments. Using metrics from social media analysis, we characterize different aspects of content dynamics such as number of shares over time, lifetime of a topic, and speed of spread, for the chosen narratives. We associate the peaks in online activity to corresponding peaks in general media for a given topic to identify the sources driving the creation and spread of misinformation. We also compute graph-theoretic measures of the graphs representing user interactions to identify key users in a narrative. We then use the empirical evidence to design a set of recommendations to minimize the spread of misinformation in social media, with the ultimate goal of providing a fast and effective response to epidemic outbreaks in the future. Supported by the empirical evidence, it becomes increasingly clear that social media platforms should incorporate regulations to minimize and prevent the spread of misinformation. These actions are vital to lessen the adverse impact on the general health of human society.