New Mathematical Approach to Geographic Profiling

• NCJ Number: 237985
• Author(s): Mike O'Leary
• Date Published: December 2009
• Length: 96 pages
• Grant Number(s): 2007-DE-BX-K005
• Agencies: NIJ-Sponsored
• Publication Type: Report (Study/Research)
• Award: The Mathematics of Geographic Profiling

Annotation

This study developed a new mathematical framework for the geographic features of a series of crimes committed by the same serial offender, and a software was developed based on the mathematical framework.

Abstract

Both the mathematical approach and the software tool are able to account for geographic features of the crime sites selected by an offender; geographic features that affect the distribution of potential anchor points (offender's home, place of work, or some other location important to the offender); differences in the travel distances of different offenders; and certain demographic characteristics (race/ethnic group, age, and sex) of the offender. The mathematical framework is described in detail in order to show how to estimate the location of an offender's anchor point by using data on the locations of the crime series; the locations of historically similar crimes; a list of solved historical crimes, with both the locations of the offense site and the corresponding anchor point; and census data, along with any available demographic information about the offender. Software was developed for implementing the mathematical algorithms. The software was built in two parts: a program called Profiler that performs all of the mathematical analysis and a program called Profiler GUI, with which the user interacts. Thus, an analyst who wishes to use the algorithm needs only to download the software package and run Profiler GUI. The user enters all of the data necessary for the analysis, and the Profiler program performs the analysis. When finished, the user has a map of the proposed search area and a map of the target attractiveness function. The authors advise that they have not yet examined the effectiveness of the tool or the mathematical algorithms it contains. Extensive figures and tables and a 61-item bibliography