Causal knowledge-based design of EDI controls: an explorative study

This paper proposes a causal knowledge-based expert system (CAKES) for modeling and designing electronic data interchange (EDI) controls. The case base of CAKES incorporates environmental, EDI control and EDI implementation factors. The primary reason for designing EDI controls is to promote consideration of the complex causal relationships that exist among various environments and EDI controls. A better understanding of industry, organization, and system characteristics is a necessary prerequisite to design, since it is difficult even for EDI experts to cognitively predict the causal effect of one control on the others. A fuzzy cognitive map is used for building a causal knowledge-base for the design. Initially, structural equation models (SEMs) are used to identify the relevant relationships among the components and to indicate their direction and strength. A casual inference in control design is possible, among environments, EDI controls and EDI implementation, through the CAKES-EDI.