U.S. flag

An official website of the United States government, Department of Justice.

NCJRS Virtual Library

The Virtual Library houses over 235,000 criminal justice resources, including all known OJP works.
Click here to search the NCJRS Virtual Library

Computer-based Retrieval of Suspects Using Similarity of Modus Operandi

NCJ Number
194886
Journal
International Journal of Police Science and Management Volume: 4 Issue: 1 Dated: Spring 2002 Pages: 5-15
Author(s)
Kaeko Yokota; Shoichi Watanabe
Date Published
2002
Length
11 pages
Annotation
This article focuses on the suspect retrieval system based on modus operandi (MO) in Japan.
Abstract
Linking crimes to a certain criminal is one of the most important tasks for police investigation. Exploring MO is sometimes an effective way to link a series of crimes to a certain perpetrator during the investigation. A MO is a summary of the habits, techniques, and peculiarities of behavior. Three premises are necessary in order to associate a series of crimes with a single prior offender based on MO: the recidivism of criminals, the differentiation of MO between offenders, and the consistency of the offender’s MO. A new computerized MO system has been developed to associate a group of crimes with a single perpetrator, based on statistical analysis. Three characteristics can be pointed out as its advantages. First, the Institute created a large database of the MO of prior offenders. Secondly, the system treats MO as a set of actions. Finally, the suspects in the database are prioritized according to MO similarities. The process of retrieving possible suspects from the database of prior offenders has two steps. The system calculates a probability for each person in the database, representing the similarity of a set of MO between each record and those recognized in a crime under the investigation. Every person in the database is then rank ordered according to the probability scores. The computer-based retrieval system was evaluated by a simulation experiment, relying on a database that stored records of a large number of prior burglary offenders. The result showed that the retrieval algorithm with the random choice method achieved satisfying accuracy. This suggests that the system has a high potential to link crimes accurately to possible offenders, prioritizing suspects according to their rank scores. 16 references