Crime Incident Association with Consideration of Narrative Information
Xiaofeng Wang, Donald E. Brown, James H. Conklin · 2007
Law enforcement agencies need to discover underlying patterns of crimes in a short time. One of the key steps is associating related criminal incidents with each other. This paper describes a new methodology to associate crime incidents automatically, as well as, accurately by using narrative information. Specifically this paper shows a new method for measuring the similarity between crime incident narratives. Evaluations show that our method could measure similarity between narrative information quickly and accurately. In addition, a toolkit based on this methodology was developed to perform crime incident association.