Time-Dynamics of (Mis)Information Spread on Social Networks: A COVID-19 Case Study

Abstract

In our study, we investigate the persistence of misinformation in social networks, focusing on the longevity of discussions related to misinformation. We employ the CoVaxxy dataset, which encompasses COVID-19 vaccine-related tweets, and classify tweets as reliable/unreliable based on non-credible sources/accounts. We construct separate networks for retweets, replies, and mentions, applying centrality metrics (degree, betweenness, closeness) to assess tweet significance. Our objective is to determine how long tweets associated with non-credible sources remain active. Our findings reveal a noteworthy correlation: tweets with longer lifespans tend to be influential nodes within the network, while shorter-lived tweets have less impact. y shedding light on the longevity of misinformation within social networks, our research contributes to a better understanding of misinformation propagation dynamics. These insights can inform strategies to combat misinformation during public health crises like the COVID-19 pandemic.