Designing and Understanding Forensic Bayesian Networks using Argumentation

Abstract

The rise of forensic evidence in court has confronted the legal domain with a number of difficulties. It appears that a communication gap may exist between forensic and legal experts.Judges, lawyers and other legal experts are accustomed to argumentative reasoning, whereas forensic experts usually quantify uncertainty with probabilities. This has resulted in a heated discussion among legal scholars about the role of numerical analyses of evidence in court. It has been argued that the source of the discussion may lie in the different ways in which experts (legal and forensic) deal with uncertainty of evidence. Argumentation theory and probability theory provide two different perspectives on uncertainty. In this thesis I combine these two perspectives in an attempt to unite the worlds of legal and forensic reasoning with uncertain legal evidence.