Using a genetic algorithm-based system for the design of EDI controls: EDIGA

The extent of major advantages and benefits derived from Electronic Data Interchange (EDI) depends upon the usage of EDI controls. EDI cannot be adopted and implemented if users are not sure about the advantages of the system. Auditors should concentrate their limited IS resources to design and effectively implement the controls that lead to highest EDI implementation and performance. This study tries to determine the Relative Importance that different modes of controls have in determining EDI implementation and performance, which will aid the design of EDI controls by auditors. This study proposes EDIGA (EDI-controls design using Genetic Algorithms), a hybrid optimization model using genetic algorithms for the design of EDI controls, that combines the search efficiency of GA with the simplicity of statistical technique, regression analysis to identify the Relative Importance of each mode of EDI controls. The estimated parameters in EDIGA are analyzed to obtain insights on the effect of EDI controls on EDI implementation and performance. The empirical investigation consists of testing EDIGA, which employs a pattern directed search mechanism to produce the best fitting power model versus linear regression analysis as a method to evaluate the effectiveness of EDIGA. It turns out that the predictive accuracy of EDIGA outperforms that of linear regression analysis. The results of study will help companies to implement EDI controls successfully. (C) 2000 Published by Elsevier Science Ltd.