Misinformation Spread and Network Centrality Features across Social Media Cascades

Abstract

The rapid proliferation of misinformation across online platforms represents a critical challenge to contemporary societal stability, democratic processes, and public health. This benchmark study investigates the structural dynamics of information diffusion by analyzing the relationship between network centrality features and the spread of unverified claims within social media cascades. Utilizing a massive dataset of distinct diffusion networks, this research isolates the topological characteristics of users who participate in the sharing of verified news versus those who propagate falsehoods. The study systematically evaluates multiple dimensions of network centrality, including degree, betweenness, closeness, and eigenvector centrality, to determine their predictive power regarding cascade depth, breadth, and virality. Findings indicate that misinformation cascades are structurally distinct, relying heavily on peripheral nodes that achieve transient high betweenness centrality, acting as critical bridges between largely disconnected echo chambers. Furthermore, the velocity of false information spread is strongly correlated with the rapid activation of nodes possessing high eigenvector centrality within the early stages of the cascade lifecycle. By mapping these structural signatures, this paper provides robust empirical evidence supporting the use of topological feature engineering for early detection systems. The insights derived from this comprehensive analysis offer profound theoretical and practical implications for platform administrators, policymakers, and researchers aiming to mitigate the impact of malicious information campaigns without compromising the principles of free expression.