Multimedia as a vehicle for knowledge modeling in expert systems

Abstract

This article examines the integration of expert systems (ES) with multimedia technology. Multimedia allows a wide variety of sensory information and knowledge to be stored in ES and thus allows ES to be more effective. Additionally, this article explores the types of knowledge that should be embedded in expert systems to provide users with support in decision-making tasks. Traditional knowledge representation schemes, such as frames and production rules, have not always proved effective for presenting requisite knowledge to the user. Traditional knowledge representations force users to translate their conceptualizations of problem attributes and characteristics into text form and then convert back to their own conceptualizations in interpreting the recommended solutions. To provide better support to users, both a knowledge representation and a knowledge presentation are necessary to foster communication, knowledge explanation, as well as user learning and corporate training. The notions of knowledge representation and knowledge presentation are derived from the concepts of symbolic and replica models proposed by Chapanis (1961). Replica models, a form of knowledge presentation, enhance communication and learning. The use of multimedia as a vehicle for knowledge presentation provides a powerful interface with a full range of audio/visual features to express knowledge more completely. This greatly enhances a corporation's ability to capture, embed, and distribute knowledge.