ANALYZING KNOWLEDGE-BASED SYSTEMS WITH MULTIVIEWPOINT CLUSTERING ANALYSIS

Knowledge-based systems have wide commercial applicability. However, a credible validation methodology for knowledge-based systems is currently lacking. Better knowledge acquisition techniques as well as better management, understanding, and enhancement of the knowledge base is critical to the success of any verification or validation activities. Our research addresses the feasibility of partitioning rule-based systems into a number of meaningful units to enhance the comprehensibility, maintainability, and reliability of expert systems software. Preliminary results have shown that no single structuring principle or abstraction hierarchy is sufficient to understand complex knowledge bases. We therefore propose the multiviewpoint clustering analysis (MVP-CA) methodology to provide multiple views of the same expert system. MVP-CA provides an ability to discover significant structures within the rule base by providing a mechanism to structure both hierarchically (from detail to abstract) and orthogonally (from different perspectives). Here we describe our approach to understanding large knowledge bases via MVP-CA. We demonstrate the need for MVP-CA by use of a couple of small classic rule bases, as well as a deployed knowledge-based system that navigates the space shuttle's reentry. We also discuss the impact of this approach on verification and validation of knowledge-based systems. MVP-CA provides art essential first step toward building an integrated environment for verification and validation of knowledge-based applications. It allows one to build reliable knowledge-based systems by suitably abstracting, structuring, and otherwise clustering the knowledge in a manner that facilitates its understanding, manipulation, testing, and utilization.