“…However, a fundamental challenge is reducing the barriers to machine understanding (Alonso, ), given even in binary classification schemes some measure of disagreement exists (Hashemi & Hall, , ). Approaches to classification tend to either be fully machine (e.g., Alghamdi & Selamat, ; Sabbah, Selamat, & Selamat, ; Scrivens, Davies, Frank, & Mei, ) or human (e.g., Droogan & Peattie, ; Ligon, Hall, & Braun, ; Torres, Jordán, & Horsburgh, ; Torres‐Soriano, ). Very few studies to date compare performance of humans and machines when classifying the underlying themes of radicalizing content.…”