Extending the Reach of Knowledge Engineering Practice beyond the Frontiers of Formal Sectors Using a Feature and Function Based Classification

Abstract

Abstract: In recent years, there had been many frameworks on improving Knowledge Engineering (KE). These frameworks had emerged for reasons ranging from shortening KE product development time, performance improvement, to addressing knowledge acquisition challenges. Despite these, impact of KE products (Expert Systems (ES) and Knowledge Based Systems (KBS)) has not been felt in non-formal sectors where there is no formalized way of keeping and exchanging their specialized knowledge. Here, heuristics and implicit tacit knowledge are often used to solve significant problems. When custodians of such specialized knowledge die, their untapped problem solving skills perish with them. Such is common in Africa, especially among native farmers, hunters and healers. Extending the reach of KE practice to these sectors can effectively help overcome this challenge. This however requires a better understanding of existing KE frameworks in a bid to isolate gaps responsible for this oversight and factor solution to this problem into subsequent design of a KE framework. To this end, this paper provides a concise feature and function based classification and review of eight prominent KE frameworks and models (CommonKADS, MIKE, MOKA, PROTÉGÉ II, SPEDE, RLM, CRLM, and PGM/CPGM) in common use in recent years and consequently recommends development of a KE framework that addresses the isolated problem.

Keywords: Expert Systems, Non-formal Sectors, Knowledge Engineering, Specialized Knowledge, Domain Knowledge Custodians