The current study addresses the shortcomings of ideology-specific radicalization models by introducing a novel framework for detecting and measuring extremist rhetoric across online communities. The authors provide a broad psychosocial model, dubbed “The Extremist Eleven,” to evaluate extremism at the user and community levels by examining linguistic behavioral characteristics. When applied to a variety of communities, such as the incel community, the technique demonstrates that extremist characteristics endure after entry and forecasts user entry with a growing degree of accuracy months in advance. The results provide a scalable, cross-ideological method for comprehending online extremism that goes beyond conventional models.
