The design of EDI controls using case-based reasoning: EDICBR

Abstract

This paper gives a description of EDICBR (EDI Controls design using Case-Based Reasoning), a case-based reasoning model for generating recommendations of EDI controls. The case base of EDICBR is composed of slots that include environmental and EDI controls. This system makes it possible to propose the EDI controls most needed in certain organizational contexts. This is performed by the following two steps. First, it suggests that the most probable level of controls from the cases be retrieved. Second, it determines the required controls that have the highest performance among the retrieved cases. EDICBR uses past cases to remind EDI auditors of similar cases and provides control recommendations. As a method to evaluate the effectiveness of EDICBR, the predictive power of the system is compared with that of multivariate discriminant analysis (MDA). The results show that the case-based reasoner outperforms MDA in predictive accuracy.