The Agent-Based Spatial Simulation to the Burglary in Beijing

摘要

Since the Agent-based simulation tool was introduced into criminology research, most work concentrated on crime theory validation or hypothesis testing, little was contributed to crime spatial pattern replication. In this paper, using street network and subway network as the landscape and proposing a statistic-based instead of predefined human mobility pattern as the individual's routine activity, the spatial distribution of burglary in Beijing is simulated and valid by the actual pattern. The result indicates that the Agent-based modeling method partly detects the crime hotspots and the spatial pattern of crime, and specifically the crime level on the nodes with different accessibility is proved to be identical to the actual one. The study made in this work demonstrates that Agent-based modeling is a potential tool to predict or explain crime pattern in space, and also some further work which aims to improve its validation is discussed in the end of this paper.