Abstract:

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.