Methods for building Conditional Probability Tables of Bayesian Belief Networks from limited judgment: An evaluation for Human Reliability Application

• Methods to build Conditional Probability Tables from limited expert judgment are evaluated.
• The focus is on Human Reliability Analysis but the conclusions can be generalized.
• Evaluation criteria: factor influences and interactions, uncertainty and scalability.
• Modelling limitations: multi-factor relationships and uncertainty modeling.

The present paper evaluates five methods for building Conditional Probability Tables (CPTs) of Bayesian Belief Networks (BBNs) from partial expert information: functional interpolation, the Elicitation BBN, the Cain calculator, Fenton et al. and Røed et al. methods. The evaluation considers application to a specific field of risk analysis, Human Reliability Analysis (HRA). The five methods are particularly suited for HRA models calculating the human error probability as a function of influencing factor assessments. The performance of the methods is evaluated on two simple examples, designed to test aspects relevant for HRA (but not exclusively): the representation of strong factor influences and interactions, the representation of uncertainty on the BBN relationships, and the method requirements as the BBN size increases. The evaluation underscores modelling limitations related to the treatment of multi-factor interdependencies and of different degrees of uncertainty in the factor relationships. The functional interpolation method is the least susceptible to these limitations; however, its elicitation requirements grow exponentially with the model size. Besides expert judgment, HRA applications of BBNs include the use of empirical data, combination of data and judgment, information from existing HRA methods: the building of the CPTs in these applications is outside the scope of the evaluation.