A prototype belief network-based expert systems shell

  • Authors:
  • Shijie Wang;Marco Valtorta

  • Affiliations:
  • Department of Computer Science, University of South Carolina, Columbia, South Carolina;Department of Computer Science, University of South Carolina, Columbia, South Carolina

  • Venue:
  • IEA/AIE '90 Proceedings of the 3rd international conference on Industrial and engineering applications of artificial intelligence and expert systems - Volume 1
  • Year:
  • 1990

Quantified Score

Hi-index 0.00

Visualization

Abstract

A belief network-based expert system works in a way entirely different from a rule-based expert system. In such systems, a high degree of non-determinism is present in the process of belief propagation. Existing belief network-based systems thus provide no control mechanism and let their users make decisions at every stage in the process of evidence gathering and belief propagation. However, for any real world expert system with large complex belief network, it is important to have efficient control of the inference procedure.In this paper we describe a novel expert system shell that incorporates a general inference control mechanism for efficient belief propagation in belief networks. The control knowledge supplied by domain experts is encoded in some action rules. With these action rules, the system can effectively direct the inference procedure, help the user to gather the most relevant evidence based on the results of previous stages, while the user of the system is still allowed to take the initiative.