A great deal has been published recently concerning expert systems and the viability of using expert system shells and a useful feature published in Systems International, April, 1988, pp 23–38, provides a summary of these shells and their market. There is, of course, great activity in all aspects of development of these systems. For example, the European experts systems market alone is said to be growing at a rate of 50 per cent a year and should reach $1 billion by 1991. In an introduction to this summary it is stated that some 40 per cent of expenditure on artificial intelligence (AI) is on software, and of this a third comes from the sale of AI languages. This feature on expert systems looks at the human and organisational problems to be considered when implementing such a system. It also outlines the market for the systems, and surveys some of the products. Finally, the limitations inherent in the rules govering expert systems shells are discussed. In an article in this review Sue Ottley, a psychologist in the human factors group at ICLs Marketing and Technical Strategy Centre, warns of some human and organisational problems to consider when implementing an expert system, from capturing an expert's knowledge to gaining user acceptance. She says that there are wide ranging concerns at the organisational level, which includes changes in the information flow patterns within an organisation, changes that can be brought about to the authority and decision‐making hierarchies and the effects of changing the locus of control and/or authority within an organisation. This latter state can have a strong effect on the acceptance of an expert system and needs to be planned for in advance. Otherwise, unplanned change or uncontrolled change, can, she maintains, be disruptive to even the most well developed and stable organisation.
MCB UP Ltd
Copyright © 1988, MCB UP Limited