KAMET II: An open diagnosis knowledge-acquisition methodology

Abstract

Different methodologies have been developed to solve various tasks such as classification, design, planning, scheduling and diagnosis. Diagnosis is a task of which the desired output is a malfunction of a system. KAMET (Knowledge-Acquisition Methodology) is a knowledge engineering methodology aimed to attack diagnosis tasks exclusively. In this article KAMET II, the second version of KAMET, will be presented with the objective of knowing its most important characteristics as well as its modeling notation which will subsequently be necessary for the knowledge bases, Problem-Solving Methods (PSMs) and the knowledge model specification. KAMET is a model-based methodology appointed to administer knowledge acquisition from multiple knowledge sources. The methodology provides a mechanism by means of which knowledge acquisition is achieved in an incremental fashion and in a cooperative environment. One important feature is the specification used to describe knowledge-based systems independently of their implementation. A four-component architecture is presented to achieve this goal and to allow component separation and consequently component reuse.