Platform Affordances and HCI Interventions for Detecting Coordinated Hate Networks on X

Social media platforms like X have become spaces that foster hate and normalize verbal violence in public discourse. This study examines hate speech on X by analyzing accounts linked to five major Spanish news outlets: El País, 20 Minutos, ABC, La Vanguardia, and El Mundo. Using 448,354 messages from the 1,000 most active users over a year, it explores hate posters’ traits, community structures, targets, and writing styles. Results show these users spread hate through short, direct messages mainly aimed at journalists, politicians, and media outlets. Their posts display stylistic similarity, low readability, and strong interconnection around highly polarized “kamikaze accounts.” This human-computer interaction (HCI) lens reveals platform affordances amplifying hate through algorithmic centrality and low readability, proposing interpretable UI designs for moderation dashboards.