Arguments, Scenarios and Probabilities: Connections Between Three Normative Frameworks for Evidential Reasoning (Abstract)

Bart Verheij, Floris J. Bex, Sjoerd Timmer, Charlotte Vlek, John-Jules Meyer, Silja Renooij, Henry Prakken

In the full paper [8], of which the present text is a compressed contribution at BNAIC 2016, we study connections between arguments, scenarios and probabilities as normative frameworks in reasoning with evidence. Such a study is relevant given the different backgrounds of the people involved in criminal investigation and decision-making: Arguments and scenarios are familiar among fact-finders and lawyers, whereas probabilities are prominent in reports by forensic experts. By studying connections between arguments, scenarios and probabilities, we aim to enhance the understanding of these three tools for evidential reasoning in the law, thereby contributing to the reduction of reasoning errors and miscommunication caused by these different backgrounds. Our work builds on recent developments to study reasoning with forensic evidence probabilistically, and in particular using Bayesian networks [5, 2]. Since it is known that it is easy to misinterpret Bayesian networks, for instance causally, we have started the exploration of the combined modelling of arguments and scenarios. Our approach continues earlier work on the design of structured probabilistic models and their explanation [4, 3]. 1The full paper of this compressed contribution at BNAIC 2016 has been published in 2016 in Law, Probability & Risk, 15, 35-70. http://dx.doi.org/10.1093/lpr/mgv013.We review research on the formal and computational connections between three normative frameworks for evidential reasoning based on arguments, scenarios and probabilities, respectively. We study pairwise connections, and connections between all three. Specifically, we discuss a hybrid model connecting arguments and scenarios, showing how reasoning with arguments can be combined with reasoning to the best explanatory scenario (see also [1]). We show how scenarios can be embedded in Bayesian networks, thereby connecting the role of the global coherence of scenarios with degrees of uncertainty (see also [9] where a case study evaluation is performed). We show how arguments can be extracted from a Bayesian network (see also [6]). We propose a view on arguments to and from scenarios in the context of probability theory (see also [7]). The paper explains the motivation for the different parts of the research and discusses strengths and weaknesses of each part. There are many remaining hard questions about the safe handling of probabilistic and non-probabilistic evidence in criminal investigation and decision-making. Still we hope that the lessons that we have learnt by studying the different connections between arguments, scenarios and probabilities, will gradually contribute to the prevention of reasoning errors, and a reduction of miscommunication between fact-finders and forensic experts.

This is an abstract of a full paper in the journal Law, Probability and Risk.

Manuscript (in PDF-format)

Reference:
Verheij, B., Bex, F.J., Timmer, S., Vlek, C., Meyer, J.J., Renooij, S., & Prakken, H. (2016). Arguments, Scenarios and Probabilities: Connections Between Three Normative Frameworks for Evidential Reasoning (Abstract). Proceedings of the 28th Benelux Conference on Artificial Intelligence (BNAIC 2016) (eds. Bredeweg, B., & Bosse, T.), 192-193.


Bart Verheij's home page - research - publications