Human factors in expert system design: can lessons in the promotion of methods be learned from commercial DP?

Abstract

The way that human factors methods and information are presented to designers of expert systems is extremely important if such methods are to be widely adopted. The nature of human factors knowledge makes such presentation problematic. The problems are somewhat analogous to those encountered in the promotion of ‘structured development standards’ within the commercial data processing environment. In commercial DP, structured development standards are promoted on the basis that increased development costs incurred at an early stage can be offset by reduced maintenance and enhancement costs later on. Within human factors, the user-centred design approach to expert system development is promoted on the basis that it will increase the likelihood of producing ‘useful’ and ‘usable’ systems. In neither DP nor human factors, can tangible benefits of adopting more rigorous methods be provided. Both, however, appeal, to some extent, to the ‘common sense’ of the system developers. By comparing the two fields the paper assesses whether human factors specialists might learn from their DP counterparts when attempting to promote the adoption of appropriate methods. One feature of DP structured development standards which is pertinent to human factors is that of ‘tailorability’. The paper discusses the suggestion that the adoption of the human factors methods will largely depend on the way that such methods can be tailored to meet the requirements of expert system development in industrial and commercial environments. An analysis of these requirements, through a series of case studies, is being conducted, and this work is briefly described.