Research On Spatio-Temporal Characteristics Of Crime And Visualization Based On Street-Network

Police patrol, one important part of public security prevention and control system in our country, plays a significant role in crime prevention and crime reduction. However, there still exists some problems in current police patrol model, such as unreasonable dispatch of police force, unscientific design of the patrol scheme, and uncertainty of the prevented objects, which have seriously affected the efficiency of the police patrol. Therefore, police patrol has become an important research focus of police departments all over the country to further enhance the capabilities of crime prevention.In order to increase the efficiency of the police patrol, the paper focuses on the burglaries occurring in Kecheng District, Quzhou city from 2011 to 2014 and takes the street network as the perspective. The principal research contents are as follows:1. Carrying on an investigation into the spatio-temporal characteristics of crime in the street-network. Utilizing the Crime Lorenz Curve and the Crime Gini Coefficient, we find that crime has the characteristic of spatial concentration. The small proportion of hot street segments account for large proportion of crime. For instance,20.08% of hot street segments account for 60.96% of crime. Furthermore, crime has the characteristic of spatio-temporal concentration. The small proportion of hot spatial-temporal street segments account for large proportion of crime. For instance,20.00% of hot spatial-temporal street segments account for 75.32% of crime. By analyzing the variability of crime of street segments and spatio-temporal street segments among three years, the study further shows that crime has the similar concentration characteristics and concentration degree in every year, and crime hotspot and coldspot have strong stability.2. Putting forward a method of identifying and visualizing the spatio-temporal hotspots of crime in the spatio-temporal street-network, namely street-network spatio-temporal kernel density estimation. Making use of the method, we are able to recognize and visualize the spatio-temporal clusters of crime in the spatio-temporal street-network. According to the map of spatio-temporal crime hotspots, patroller can accurately quickly readily find where the hot spatial-temporal street segments are.According to the characteristics of crime spatial-temporal on street-network and the map of hot spatio-temporal street segment, police department can scientifically dispatch patrol force, plan the scheme of patrol, and precisely put the patrol force into the time and space in which crime will occur mostly, thereby further improving the efficiency of the police patrol.