Forschungsprojekt: Data Mining for Crime Information

Global Network: Heidelberg – University College London – Louisiana State University – Temple University Philadelphia: “Data Mining of Crime Information from the General Public”

Crime in all its facets is an integral part of our daily life and affects society. Hence, crime protection and crime combat by means of geospatial technology and geographic profiling methodologies have gained increasing importance and supplement and enhance traditional criminological modus operandi for tactical and strategic decision-making. However, its full potential has not been exploited so far and current techniques show some limitations. In particular, current geospatial technology are limited in handling the constantly increasing large amounts of data and do not consider voluntarily provided subsidiary data sources (e.g., the general public, information, and hints of eyewitnesses).

Thus, the objective of this project to promote data mining methods, particularly the self-organizing map algorithm, to explore hidden information in such data sources provided by the general public. This allows the analysis of a priori not readily apparent patterns and may gain insight into the complex behavior of crime. Furthermore, the proposed framework is evaluated by two case studies, both located in the US.

The present proposal seeks to initiate an international research exchange on crime modeling by means of geographic information systems and data mining techniques. To tackle the expected challenges is best addressed in an interdisciplinary and international research community of leading experts, allowing for knowledge transfer, inducing spillover effects and establish the University of Heidelberg as a future competence center for crime data mining.