KIRAS Security Research

2012

Crime Predictive Analytics (CriPA)

Crime Predictive Analytics is concerned with the predictive analysis of crime and includes quantitative methods to identify patterns and dependences in data bases. Thus, future events may be predicted and potential strategic decisions may be evaluated

Reliable predictions and risk analysis of short- and long-term trends of crime are valuable tools for strategic crime prevention for the Austrian law enforcement management.

Crime Predictive Analytics is concerned with the predictive analysis of crime and includes quantitative methods to identify patterns and dependences in data bases. Thus, future events may be predicted and potential strategic decisions may be evaluated.

In the present project, appropriate methods and software components are to be developed and integrated in a reference system, for two important questions.

The first involves long-term, large-scale predictions to estimate future trends of crime for the decision support of strategic measures to crime prevention. As social changes in demographics and the employment structure considerably influence crime, effects of these changes on crime are analyzed. These results and further factors of influence are used for prediction of medium- and long-term trends of crime.

For short-time, small-scale predictions and risk models, prevention is most important. Small-scale predictions of the occurrence of crime and an appropriate integration into a geographic information system (GIS) support the resource planning and strategic crime prevention of the law enforcement management. Spatial models for risk analysis are developed, implemented and validated. In doing so the early identification of spatial and temporal patterns as well as the consideration of relevant factors of influence on crime and the integration of these methods into a GIS is very important.

Evaluations and experiences of police experts are systematically recorded, prepared and integrated into the models to improve the precision of the predictions.

All models are implemented in a real-time capable, web-based, GIS-based reference system. The validation of the system is based on real data.

The integration of the developed Predictive Analytics methods into the dashboard of the Austrian law enforcement management is intended.

Project leader
JOANNEUM RESEARCH Forschungsgesellschaft mbH, DI Ulrike Kleb

Project partner
Bundesministerium für Inneres, Bundeskriminalamt
Paris-Lodron Universität Salzburg, Interfakultärer Fachbereich für Geoinformatik – Z_GIS
Institut für Rechts- und Kriminalsoziologie
SynerGIS Informationssysteme GmbH

Contact
DI Ulrike Kleb
JOANNEUM RESEARCH Forschungsgesellschaft mbHPOLICIES – Zentrum für Wirtschafts- und Innovationsforschung
Leonhardstraße 59
8010 Graz
tel.: +43 316 876-1555
fax: +43 316 8769-1555
E-mail: ulrike.kleb@joanneum.at
web: www.joanneum.at/policies

print