A Representation Scheme for Managing Complex Professional Knowledge

Abstract

A representation scheme called CKR-1 is introduced to deal with the challenges of representing complex knowledge.  Complex knowledge is defined as deep knowledge concerning a complex object, event, situation or process.  The nature of complex knowledge is identified from samples of professional knowledge drawn from the insurance industry.  These examples highlight efficacy of the scheme as well as its limitations.  CKR-1 is a type of semantic network that supports simple and abstract concepts, events, activities and situations.  It also supports several types of relationships including business, logical, causal, process etc.  It derives its expressivity from ability to support abstractions, elaborations, assertions and alternative points of view.  Usability was tested with a randomized selection of fifty knowledge items from a book for insurance professionals.  The scheme handled 39 of these well.  Despite this, the scheme suffers from limitations stemming from the inherent difficulty in expressing concepts involving several other vaguely defined abstract concepts.