Analyzing the Simonshaven Case Using Bayesian Networks

This paper is one in a series of analyses of the Dutch Simonshaven murder case, each using a different modeling approach. We adopted a Bayesian network (BN)-based approach which requires us to determine the relevant hypotheses and evidence in the case and their relationships (captured as a directed acyclic graph) along with explicit prior conditional probabilities. This means that both the graph structure and probabilities had to be defined using subjective judgments about the causal, and other, connections between variables and the strength and nature of the evidence. Determining if a useful BN could be quickly constructed by a small group using the previously established idioms-based approach which provides a generic method for translating legal cases into BNs, was a key aim. The model described was built by the authors during a workshop dedicated to the case at the Isaac Newton Institute, Cambridge, in September 2016. The total effort involved was approximately 26 h (i.e., an average of 6 h per author). With the basic assumptions described in the paper, the posterior probability of guilt once all the evidence is entered is 74%. The paper describes a formal evaluation of the model, using sensitivity analysis, to determine how robust the model conclusions are to key subjective prior probabilities over a full range of what may be deemed "reasonable" from both defense and prosecution perspectives. The results show that the model is reasonably robust-pointing not only generally to a reasonably high posterior probability of guilt but also generally below the 95% threshold expected in criminal law. Given the constraints on building a complex model so quickly, there are inevitably weaknesses; hence, the paper describes these and how they might be addressed, including how to take account of supplementary case information not known at the time of the workshop.

[1]  Henry Prakken,et al.  Arguments, scenarios and probabilities: connections between three normative frameworks for evidential reasoning , 2016 .

[2]  J. Koehler Error and Exaggeration in the Presentation of DNA Evidence at Trial , 1993 .

[3]  A Gammerman,et al.  Statistical modelling in specific case analysis. , 1996, Science & justice : journal of the Forensic Science Society.

[4]  A. Dawid,et al.  Forensic identification with imperfect evidence , 1998 .

[5]  Norman Fenton,et al.  Avoiding probabilistic reasoning fallacies in legal practice using Bayesian networks , 2011 .

[6]  Amanda B. Hepler,et al.  Object-Oriented Graphical Representations of Complex Patterns of Evidence , 2007 .

[7]  Martin Neil,et al.  Inference in hybrid Bayesian networks using dynamic discretization , 2007, Stat. Comput..

[8]  Franco Taroni,et al.  How the probability of a false positive affects the value of DNA evidence. , 2003, Journal of forensic sciences.

[9]  Marek J. Druzdzel,et al.  Impact of precision of Bayesian network parameters on accuracy of medical diagnostic systems , 2013, Artif. Intell. Medicine.

[10]  William Marsh,et al.  How to model mutually exclusive events based on independent causal pathways in Bayesian network models , 2016, Knowl. Based Syst..

[11]  Henry Prakken,et al.  A method for explaining Bayesian networks for legal evidence with scenarios , 2016, Artificial Intelligence and Law.

[12]  Christian Jowett Lies, Damned Lies, and DNA Statistics: DNA Match Testing, Bayes' Theorem, and the Criminal Courts , 2001, Medicine, science, and the law.

[13]  Judea Pearl,et al.  Probabilistic reasoning in intelligent systems , 1988 .

[14]  L. C. van der Gaag,et al.  Building probabilistic networks: Where do the numbers come from? - a guide to the literature , 2000 .

[15]  Floris Bex,et al.  Arguments, Stories and Criminal Evidence - A Formal Hybrid Theory , 2011, Law and philosophy library.

[16]  Henry Prakken,et al.  Building Bayesian networks for legal evidence with narratives: a case study evaluation , 2014, Artificial Intelligence and Law.

[17]  Anne S. Hsu,et al.  When 'neutral' evidence still has probative value (with implications from the Barry George Case). , 2014, Science & justice : journal of the Forensic Science Society.

[18]  Henry Prakken,et al.  A hybrid formal theory of arguments, stories and criminal evidence , 2010, Artificial Intelligence and Law.

[19]  David Lagnado,et al.  Modelling competing legal arguments using Bayesian model comparison and averaging , 2019, Artificial Intelligence and Law.

[20]  A. Philip Dawid,et al.  USING A GRAPHICAL METHOD TO ASSIST THE EVALUATION OF COMPLICATED PATTERNS OF EVIDENCE , 1997 .

[21]  Henry Prakken,et al.  An abstract framework for argumentation with structured arguments , 2010, Argument Comput..

[22]  Norman Fenton,et al.  Risk Assessment and Decision Analysis with Bayesian Networks , 2012 .

[23]  David A. Lagnado,et al.  A General Structure for Legal Arguments About Evidence Using Bayesian Networks , 2013, Cogn. Sci..

[24]  N. Fenton A General Structure for Legal Arguments Using Bayesian Networks , 2010 .

[25]  D. Schum,et al.  A Probabilistic Analysis of the Sacco and Vanzetti Evidence , 1996 .

[26]  Colin Aitken,et al.  Bayesian Networks for Probabilistic Inference and Decision Analysis in Forensic Science , 2014 .

[27]  Linda C. van der Gaag,et al.  Making Sensitivity Analysis Computationally Efficient , 2000, UAI.

[28]  Norman Fenton,et al.  Calculating and understanding the value of any type of match evidence when there are potential testing errors , 2014, Artificial Intelligence and Law.

[29]  David J. Balding,et al.  Bayesian Networks and Probabilistic Inference in Forensic Science , 2011 .

[30]  J. Pearl,et al.  The Book of Why: The New Science of Cause and Effect , 2018 .

[31]  Kathryn B. Laskey Sensitivity analysis for probability assessments in Bayesian networks , 1995, IEEE Trans. Syst. Man Cybern..

[32]  David A. Lagnado,et al.  Legal idioms: a framework for evidential reasoning , 2013, Argument Comput..

[33]  P.E.M. Huijgen Use of Bayesian belief networks in legal reasoning , 2002 .

[34]  David A. Lagnado,et al.  The opportunity prior: a simple and practical solution to the prior probability problem for legal cases , 2017, ICAIL.

[35]  I. Evett,et al.  A hierarchy of propositions: deciding which level to address in casework , 1998 .

[36]  Norman Fenton,et al.  Bayes and the Law. , 2016, Annual review of statistics and its application.

[37]  David A. Schum,et al.  Knowledge, probability, and credibility , 1989 .