In this work, the authors developed a fundamentally new mathematical framework for the geographic profiling problem.
The primary question in geographic profiling is, given the locations of a series of crimes committed by the same serial offender, to estimate the location of that offender’s anchor point. Currently, there are three main approaches to the problem, exemplified by the three software systems- CrimeStat, Dragnet, and Rigel. Though the details of the approaches taken by these software packages differ, they share a common mathematical heritage. The authors developed a fundamentally new mathematical framework for the geographic profiling problem. Our mathematical approach to this problem is based on Bayesian inference and begins with the explicit ansatz that the offender’s choice of targets depends only on (1) the distance between the target and the offender’s anchor point, and (2) local geographic features of the target location. The algorithm requires a representative list of historical crimes of the same type as the series; this is used to estimate the local target attractiveness. This mathematical model has been implemented in software. (Published abstract provided)
Downloads
Similar Publications
- Assessment of the Effectiveness of Emergency Lighting, Retroreflective Markings, and Paint Color on Policing and Law Enforcement Safety
- Online Commercial Sexual Exploitation of Children in a National Victim Survey
- Occupational Prestige of Law Enforcement Officers: Quantifying Self and Public Perceptions of Prestige