CriPA Crime Predictive Analytics
Is it possible to predict crime?
The project CRIPA demonstrates how well this works and which methods are suitable.
Reliable predictions and risk assessments related to spatial and temporal developments in crime are indispensable strategic tools that can be used to combat crime. Predictive analytical methods that are suitable for this purpose are explored in CRIPA. In order to make long-term, large-scale predictions about developments in crime, trend models such as generalized additive models are used. Important influential factors on crime and social changes such as aging, migration or employment structures must be taken into account in these models. Short-term, small-scale predictive and risk models (e.g., risk terrain and near-repeat modeling), taken together with a corresponding representation in a geographic information system, allow the early identification of space-time related patterns and help the police form prevention measures.
In order to improve the accuracy of predictions, assessments and experiences of police field experts will be systematically collected, processed and integrated into the models. The proposed models are implemented in a real-time, web- and GIS-based reference system.
- JOANNEUM RESEARCH – DIGITAL
- Federal Ministry of the Interior
- University of Salzburg, Interfaculty Departement of Geoinformatics – Z_GIS
- SynerGIS Informationssysteme GmbH
- Institute für Legal and Criminal Sociology