行为轨迹大数据中隐藏着具有强时空相关性的时空聚类模式，蕴含着人们丰富的行为模式和活动规律。行为轨迹大数据的高性能时空聚类与社会分析是地理信息科学与工程领域迫切需要解决的关键科学问题。对行为轨迹大数据的时空聚类及其高性能求解方法进行了探讨，分析了行为轨迹时空聚类的3个应用：热点区域提取、异常轨迹探测和交通拥堵分析。这些应用分析可以为城市交通管理、社会管理以及人们的日常出行等提供有有效的参考。

The big spatiotemporal behavior trajectory data contain rich human behavior patterns and rules, and hide some spatiotemporal clustering patterns with strong spatiotemporal correlations. The high performance spatiotemporal clustering and social analysis is a key scientific research problem need to be solved in the field of geographic information science and engineering. We research the methods of spatiotemporal clustering of behavior trajectory data, and furtherly study the high performance computingproblems of t he spatiotemporal clustering, and finally analyze three applications based on these spatiotemporal clustering methods, including hotspots extraction, anomalous trajectory detection and traffic congestion analysis. These applications can provide an effective reference for urban traffic management, social management, and human daily travel.