Spatio-temporal recognition of criminal patterns: A spatial statistical approach

Crime control and prevention is related to the understanding of the spatio-temporal behavior of crimes and complex local dynamics. Therefore, this study aims to reveal patterns in the spatial distribution of property crimes over time, in terms of similarity of location and spatial concentration in relation to neighboring areas. In addition, we sought to explore the socio-economic characteristics and points of interest that increase the susceptibility to these crimes. The study was applied in the Metropolitan Region of Recife supported by spatial statistics. The critical regions are mostly in city central areas, and this is a pattern of similarity and spatial clusters that is repeated for different time frames. Crime is strongly associated with the environment, due to the presence of places of commercial interest and bus stops. The results can be used to assist in guiding and planning public security actions.