Refining a Heuristic for Constructing Bayesian Networks from Structured Arguments

Abstract

Recently, a heuristic was proposed for constructing Bayesian networks (BNs) from structured arguments. This heuristic helps domain experts who are accustomed to argumentation to transform their reasoning into a BN and subsequently weigh their case evidence in a probabilistic manner. While the underlying undirected graph of the BN is automatically constructed by following the heuristic, the arc directions are to be set manually by a BN engineer in consultation with the domain expert. As the knowledge elicitation involved is known to be time-consuming, it is of value to (partly) automate this step. We propose a refinement of the heuristic to this end, which specifies the directions in which arcs are to be set given specific conditions on structured arguments.

Deductive and abductive argumentation based on information graphs

“…1. This interpretation is based on a previous interpretation of this Wigmore chart as a preliminary version of an IG in which only causal and evidential information is considered and the roles of generalisation and inference are not separated [37]. For every claim p in the Wigmore chart, a proposition node p is included in P. As noted by Kadane and Schum [18, p. 88], the generalisations used in the inferences from the testimonies are evidential.…”

Bridging the Gap Between Informal Sense-Making Tools and Formal Systems

“…Given that the two main formal AI systems under consideration in this thesis are argumentation frameworks and BNs, we study how argumentation techniques can be used to facilitate BN construction. In previous work Wieten et al, 2018a;Timmer et al, 2015], the problem of constructing BNs from information specified in arguments about the domain was investigated (see also Section 8.2.2). In this thesis, we consider a different approach and investigate how argumentation can be used to argue about the BN under construction instead of about the domain:…”

“…In Chapter 3 we motivate and present the IG-formalism. The idea of a graph-based intermediary formalism between analyses performed using informal sense-making tools and formal AI systems was first considered by us in [Wieten et al, 2018b[Wieten et al, , 2019b in the form of so-called 'argument graphs'. Based on this earlier work the IG-formalism was introduced in [Wieten et al, 2020[Wieten et al, , 2021a], which we extended to increase its expressivity in [Wieten et al, 2021b].…”

“…In Chapter 5 BN construction is considered. The idea of constructing BNs by exploiting knowledge captured in argument graphs was introduced in [Wieten et al, 2018b], where in [Wieten et al, 2019b] the scope of this approach was extended and formal properties were preliminarily investigated. In [Wieten et al, 2021a] IGs were taken as a starting point, where a structured approach for automatically constructing a BN graph from an IG was proposed and its formal properties were investigated.…”

“…In Chapter 6 we illustrate the approaches of Chapters 4 and 5 by applying them to parts of an actual legal case, namely the well-known Sacco and Vanzetti case. We then compare the results obtained by applying the BN construction approach from Chapter 5 to a BN modelling by Kadane and Schum [1996]; this part of the case study is based on a precursory case study performed in [Wieten et al, 2018b] and serves as a preliminary validation of our approach of Chapter 5.…”