BUILDING LARGE-SCALE BAYESIAN NETWORKS
Dr Martin Neil, Centre for Software Reliability, City University and Agena Ltd, London, UK

Abstract: The benefits of using Bayesian Networks (BNs) to model uncertain domains are well known, especially since the recent breakthroughs in algorithms and tools to implement them. However, there have been serious problems for practitioners trying to use BNs to solve realistic problems. This is because, although the tools make it possible to execute large-scale BNs efficiently, there have been no guidelines on building BNs. Specifically, practitioners face two significant barriers. The first barrier is that of specifying the graph structure such that it is a sensible model of the types of reasoning being applied. The second barrier is that of eliciting the conditional probability values, from a domain expert, for a graph containing many combinations of nodes, where each may have a large number of discrete or continuous values. We have tackled both of these practical problems in recent research projects and have produced partial solutions for both that have been applied extensively on a number of real-life applications. In my presentation I will concentrate on this first problem, that of specifying a sensible BN graph structure. The proposed solution is based on the notion of generally applicable 'building blocks', called idioms, which can be combined together into modular sub-nets. These can then in turn be combined into larger BNs, using simple combination rules and by exploiting recent ideas on modular and Object Oriented BNs (OOBNs). This approach, which has been implemented in a BN tool, can be applied in many problem domains. I use examples to illustrate how it has been applied to build large-scale BNs for predicting software safety. The work presents a major breakthrough for those building BN applications.

This seminar was held at the Department of Computer Science, Royal Holloway, University of London on 14 October 1999.

back