Implementing Evidential Reasoning in Expert Systems

TL;DR

This paper presents GERTIS, a prototype expert system implementing extended Dempster-Shafer theory for evidential reasoning, which enhances explanation generation and knowledge representation in rule-based systems.

Contribution

It demonstrates the feasibility of rule-based evidential reasoning systems and introduces methods for improved explanations and richer knowledge representation.

Findings

GERTIS effectively diagnoses rheumatoid arthritis.

Two key knowledge types improve explanation quality.

The system shows enhanced knowledge representation.

Abstract

The Dempster-Shafer theory has been extended recently for its application to expert systems. However, implementing the extended D-S reasoning model in rule-based systems greatly complicates the task of generating informative explanations. By implementing GERTIS, a prototype system for diagnosing rheumatoid arthritis, we show that two kinds of knowledge are essential for explanation generation: (l) taxonomic class relationships between hypotheses and (2) pointers to the rules that significantly contribute to belief in the hypothesis. As a result, the knowledge represented in GERTIS is richer and more complex than that of conventional rule-based systems. GERTIS not only demonstrates the feasibility of rule-based evidential-reasoning systems, but also suggests ways to generate better explanations, and to explicitly represent various useful relationships among hypotheses and rules.