A Decision Support System for Forensic Entomology

This paper presents a multiagent-based model of insect development on a dead body and a three layers Decision Support System architecture able to perform retrodictive (abductive) reasoning from multiagent-based models or more generally, complex systems models. This architecture is used in order to compute post-mortem intervals from entomological data sampled on cadavers. Knowing the exact time of a death is fundamental in criminal investigations. Thus, it is necessary for experts to guarantee the reliability of their results. We show that post-mortem interval estimated with traditional entomological methods can lead to important overestimations. Indeed, these methods do not take into account all the interdependent processes involved in the development of insects such as fly population dynamics in the ecosystem or the gregarious behavior of insect larvae that can lead to local temperature increases on the body. Forensic Entomology is widely used in several countries; it is then important to develop new methodologies and tools to improve the efficiency and reliability of entomological expertises. ForenSeek – the implementation of the model and Decision Support System architecture – aims to be a software program that can be used as a post-mortem interval estimation tool and as a virtual laboratory, to simulate colonization and development cases.