This study uses quantitative methods to assess and categorize the risk of terrorist attacks, using data from the Global Terrorism Database (GTD) from 1970 to 2020. Twenty-two indicators were selected, and moment estimation theory along with four comprehensive evaluation models (linear weighted evaluation, fuzzy comprehensive evaluation, TOPSIS, and PSO-PPE) were used to identify the top 10 riskiest attacks globally. Five clustering analysis methods (FCM, CURE, DBSCAN, CLIQUE, and GMM) and three evaluation criteria were applied for risk categorization, with visual analysis using kernel density estimation. The study identified the September 11 attacks as the riskiest, and spatial analysis revealed four “turbulent cores” of risk in Central Asia, the Middle East & North Africa, South Asia, and Central America & Caribbean. Insights and recommendations for counter-terrorism efforts are provided.

Zonghuang Xu, Yao Lin, Hongyu Cai, Wei Zhang, Jin Shi, Lingyun Situ