Human versus machine: a comparison of the accuracy of geographic profiling methods

Abstract

The current article addresses the continuing debate as to the accuracy of geographic profiling methods in predicting the location of serial crime offenders. In particular, the research addresses whether spatial distribution methods and humans using simple heuristics are as accurate at predicting home locations of offenders as more complex algorithm-based methods such as RIGEL, DRAGNET, and Crimestat. Using a random sample of solved serial crimes, the research found that predictions by human judges and spatial distribution methods performed equally as well as more complex algorithm-based methods. In addition to a discussion of findings, implications for police investigations and researchers are included. Copyright © 2006 John Wiley & Sons, Ltd.