We propose to extend the current binary understanding of terrorism (versus non-terrorism) with a Dynamic Matrix of Extremisms and Terrorism (DMET). DMET considers the whole ecosystem of content and actors that can contribute to a continuum of extremism (e.g., right-wing, left-wing, religious, separatist, single-issue). It organizes levels of extremisms by varying degrees of ideological engagement and the presence of violence identified (e.g., partisan, fringe, violent extremism, terrorism) based on cognitive and behavioral cues and group dynamics. DMET is globally applicable due to its comprehensive conceptualization of the levels of extremisms. It is also dynamic, enabling iterative mapping with the region- and time-specific classifications of extremist actors. Once global actors recognize DMET types and their distinct characteristics, they can comprehensively analyze the profiles of extremist actors (e.g., individuals, groups, movements), track these respective actors and their activities (e.g., social media content) over time, and launch targeted counter activities (e.g. de-platforming, content moderation, or redirects to targeted CVE narratives).

翻译：我们提出以极端主义与恐怖主义动态矩阵（DMET）扩展当前对恐怖主义（与非恐怖主义）的二元理解。DMET考察了可能构成极端主义连续谱（如右翼、左翼、宗教、分裂主义、单一议题）的整个生态系统中的内容与行为体。它基于认知和行为线索以及群体动力，按照意识形态参与程度和识别出的暴力呈现（如党派、边缘群体、暴力极端主义、恐怖主义），对极端主义的不同层级进行组织。由于对极端主义层级进行了全面的概念化，DMET具有全球适用性。同时它也是动态的，能够根据区域和时间特定分类对极端主义行为体进行迭代映射。一旦全球行为体识别出DMET类型及其独特特征，他们就能全面分析极端主义行为体（如个人、团体、运动）的概况，追踪这些行为体及其活动（如社交媒体内容）随时间的变化，并开展针对性的反制活动（如去平台化、内容审核或定向推送反暴力极端主义叙事）。