J. Keppens and J. Zeleznikow (UK)
decision support system, crime investigation, artificial intelligence, model-based reasoning
Crime investigation is a complex task involving vast amounts of information and requiring many different types of expert knowledge. Crime investigators would therefore benefit from the use of decision support systems to help manage this information and to provide knowledge to help solve the more complex problems. Current research efforts in this area have focussed on the information management side of the problem and tend to steer clear of formalising expert knowledge. This is understandable since conven tional knowledge based systems lack the robustness needed to cope with the variety of circumstances that can be en countered during criminal investigation. However, similar problems have been encountered in the physical systems domain and were tackled by means of novel model-based reasoning techniques. This paper explores the use of robust model-based reasoning approaches to model expert knowl edge for crime investigation and it presents a framework for such systems. The preliminary ideas presented in this paper are illustrated by means of practical examples pro duced during ongoing work in the development of a system for differentiating between homicidal, suicidal, accidental and natural deaths.
Important Links:
Go Back
