Spatial estimation of public security efficiency for Beijing subdistricts based on big data

As urbanization levels continue to rise, urban public security has gradually become an important component of urban governance. This study focuses on 64 central subdistricts of Beijing and integrates multi-modal big spatial data for simulating public security costs and evaluating public security efficiency. A gradient boosting regression tree method is employed to simulate public security cost, which achieves the best fitting performance under complex geographic big data conditions, and an analytic hierarchy process method is used to evaluate the public security efficiency, combining weights of four types of security cases, i.e., theft, serious criminal cases, robbery, and cases of provoking trouble, to produce a quantitative public security effect score. Furthermore, this research proposes a dual evaluation method for public security costs and efficiency, and identifies typical subdistricts characterized by “high expenditure and low effect” (HL), “low expenditure and high effect” (LH) and “low expenditure and low effect” (LL). Experiments indicate that our method produces the best security cost simulation result compared to multiple linear regression and principal component analysis methods, and the HL subdistricts primarily correspond to high-value financial hubs (e.g., Jinrongjie) or areas in a costly transitional phase toward smart policing (e.g., Hepingjie), where additional public security investments exhibit diminishing marginal returns. In contrast, the LH subdistricts achieve high security performance through proactive daily patrols, precise risk classification, and resident self-governance, demonstrating that efficiency improvements do not necessitate increased fiscal input but rather rely on institutionalized, data-driven, and participatory management. The research does not only provide empirical evidence for the optimal allocation of public security resources but also offers reference value for refined public security governance in other major cities.