The impact of problem domains and knowledge acquisition techniques: a content analysis of P/OM expert system case studies

Abstract In the past several decades, many expert system applications have been developed and reported in the literature. Case studies of these applications typically include a detailed description of the problem domain, knowledge acquisition techniques used and also some indication of the relative success of the application. This paper presents the results of an extensive content analysis of more than 90 expert system applications in the field of production and operations management. The content analysis presented here focuses on describing the knowledge acquisition techniques used and also on the problem domain that the applications address. For years developers have speculated that certain types of problems such as planning, are more difficult, but may yield a higher-impact system. Likewise, developers and empirical KA researchers have tried to determine the conditions under which certain KA techniques will work better than others. Researchers have also speculated for years about which techniques might work best for different problem domains. By carefully analyzing this large body of case studies and operationalizing the notion of the ‘impact’ of the expert system application, we can now begin to make normative conclusions about which techniques and which problem domains seem to yield applications that have the highest impact on the respective organization. This is also the first study that will offer more tangible evidence regarding the possible linkages between problem domains and KA techniques in a more pragmatic manner than has been done previously. In an environment of shrinking IT budgets, the results of this study can provide managers with real guidance about what projects may have the biggest potential impact on their organization.