The importance of combining spatial and temporal aspects has been increasingly recognized over recent years, yet pertinent pattern analysis methods in place-based crime research still need further development to explicitly indicate spatial-temporal localities of pertinent factors’ influence ranges. This paper proposes an approach, Spatial-Temporal Indication of Crime Association (STICA), to facilitate identifying the main contributing factors of crime, which are operated at diverse spatial-temporal scales. The method’s rationale is to progressively discern the spatial zones with diverse temporal crime patterns. A specific implementation of the STICA approach, by combining kernel density estimation, k-median-centers clustering, and thematic mapping, is applied to understand the burglary in an urban peninsula, China. The empirical findings include: (1) both the main time-stable and time-varying factors of crime can be indicated with the disparities of temporal crime patterns for different spatial zones based on the STICA results. (2) The spatial range of these factors can enlighten the understanding of interactions for generating crime patterns, especially with regards to how temporally transient and spatially global factors can produce a locally crime-ridden zone through the mediation of stable factors. (3) The STICA results can reveal the spatially contextual effects of stable factors, which are of great value to improve modeling crime patterns. As demonstrated, the STICA approach is effective in exploring contributing factors of crime and has shown great potential for providing a new vision in place-based crime research.
Kernel density weighted loess normalization improves the performance of detection within asymmetrical data