Abstract: As one of the hotspots of current research, spatiotemporal data mining provides new ideas and technical means for urban research. In this paper we take Wuxi city as an example, and build the corresponding data mining model by using multi-source heterogeneous spatiotemporal data such as the mobile signaling, traffic jam data, POI data and urban planning achievement. Then we do some analysis of urban main characteristics such as human occupation and residence balance, regional contacting degree to build the urban evaluation system based on multi-source data mining, explore the application of spatiotemporal data mining in urban physical examination, in order to provide scientific evaluation and decision-making for urban planning, operation management.

Key words: urban planning evaluation, spatio-temporal data mining, mobile signaling data, multi-source data, heterogeneous