On the Advancement of Using Data Mining for Crime Situation Recognition

Abstract

The primary role of intelligence organisations and public safety agencies encompasses protecting the lives and property of citizens. However, the urban population growth rate tends to overshadow the available security resources. Thus, the security agencies appear to be more reactive than proactive. Public safety agencies usually have a plethora of under-utilised crime incident reports at their disposal, which if efficiently analysed could reveal some previously unknown useful information. Such information reveals insights into a range of functions in a crime investigation, which can assist in determining criminal trends and in knowledge-driven decision support. This chapter provides an overview of data mining techniques and existing applications used in this domain of interest. Features of existing applications and techniques, such as exploratory basis, model selection, algorithm advancement and result summary, are compared. Future potential of crime data mining, and open research issues, are also discussed.