Knowledge Engineering for Large Belief Networks

TL;DR

This paper introduces techniques for constructing and visualizing large belief networks, including the noisyMAX model, leak probabilities, and a dynamic visualization tool called Netview, to aid knowledge engineers.

Contribution

It presents novel methods for modeling causal independence with noisyMAX, uses leak probabilities for the closed-world assumption, and introduces Netview for dynamic network visualization and editing.

Findings

Effective modeling of causal independence with noisyMAX.

Leak probabilities help enforce the closed-world assumption.

Netview enables dynamic visualization and version control of belief networks.

Abstract

We present several techniques for knowledge engineering of large belief networks (BNs) based on the our experiences with a network derived from a large medical knowledge base. The noisyMAX, a generalization of the noisy-OR gate, is used to model causal in dependence in a BN with multi-valued variables. We describe the use of leak probabilities to enforce the closed-world assumption in our model. We present Netview, a visualization tool based on causal independence and the use of leak probabilities. The Netview software allows knowledge engineers to dynamically view sub-networks for knowledge engineering, and it provides version control for editing a BN. Netview generates sub-networks in which leak probabilities are dynamically updated to reflect the missing portions of the network.