Identifying "Hot Link" Between Crime and Crime-Related Locations

Abstract

Crime is typically a multi-location event where multiple locations are associated through one crime incident. Understanding the patterns of the spatial association between crime locations and their corresponding crime-related locations (for example, the residence location of an offender or that of a victim) can enhance our capability to explain and predict crime patterns. GIS technologies coupled with spatial statistics have been widely used to model areas of high crime (that is, crime hot spot). But very limited effort has been spent to investigate the spatial association between crime locations from a crime hot spot and the corresponding crime-related locations. This chapter introduces the concept of “hot link” to describe the spatial autocorrelation of the one-to-one relationship between a crime location and a crime-related location. It develops an approach for the hot link analysis problem. Related techniques are applied to the hot link analysis between auto theft and recovery locations in the city of Buffalo, New York. Identifying the patterns of hot links from auto theft locations to recovery locations is important for understanding auto thieves’ travel behavior. Related findings can contribute to the law enforcement’s effort to fight against auto theft. The hot link analysis method can be easily extended to analyzing spatial patterns of other types of crime and crime-related locations.