Crime prediction is crucial for sustainable urban development and protecting citizens’ quality of life. However, there exist some challenges in this regard. First, the spatio-temporal correlations in crime data are relatively complex and are heterogenous in time and space, hence it is difficult to model the spatio-temporal correlation in crime data adequately. Second, crime prediction at fine spatial temporal scales can be applied to micro patrol command; however, crime data are sparse in both time and space, making crime prediction very challenging. To overcome these challenges, based on the deep spatio-temporal 3D convolutional neural networks (ST-3DNet), we devise an improved ST-3DNet framework for crime prediction at fine spatial temporal scales (ST3DNetCrime). The framework utilizes diurnal periodic integral mapping to solve the problem of sparse and irregular crime data at fine spatial temporal scales. ST3DNetCrime can, respectively, capture the spatio-temporal correlations of recent crime data, near historical crime data and distant historical crime data as well as describe the difference in the correlations’ contributions in space. Extensive experiments on real-world datasets from Los Angeles demonstrated that the proposed ST3DNetCrime framework has better prediction performance and enhanced robustness compared with baseline methods. In additon, we verify that each component of ST3DNetCrime is helpful in improving prediction performance.

Theft is an inevitable problem in the context of urbanization and poses a challenge to people’s lives and social stability. The study of theft and criminal behavior using spatiotemporal, big, demographic, and neighborhood data is important for guiding security prevention and control. In this study, we analyzed the theft frequency and location characteristics of the study area through mathematical statistics and hot spot analysis methods to discover the spatiotemporal divergence characteristics of theft in the study area during the pre-COVID-19 and COVID-19 periods. We detected the spatial variation pattern of the regression coefficients of the local areas of thefts in Haining City by modeling the influencing factors using the geographically weighted regression (GWR) analysis method. The results explained the relationship between theft and the influencing factors and showed that the regression coefficients had both positive and negative values in the pre-COVID-19 and COVID-19 periods, indicating that the spatial distribution of theft in urban areas of Haining City was not smooth. Factors related to life and work indicated densely populated areas had increased theft, and theft was negatively correlated with factors related to COVID-19. The other influencing factors were different in terms of their spatial distributions. Therefore, in terms of police prevention and control, video surveillance and police patrols need to be deployed in a focused manner to increase their inhibiting effect on theft according to the different effects of influencing factors during the pre-COVID-19 and COVID-19 periods.