A large-scale data management and application analysis based on advanced classifier computing for the ERP system selection and adoption

Abstract

Enterprise resource planning (ERP), promising trend of emerged large-scale data management, has urgent needs to enterprises that are faced with competitions under external environment and globalisation trend. It is an interesting issue to help ERP system vendor selecting a suitable customer through intelligent models. This motivates the study. We compare the empirical results of the decisional feature database constructed by two classification models, Models 1 and 2, and find out the critical factors for ERP system selection summarised from the analytical results and hypothesis. The empirical results include: 1) Model 1: the accuracy of percentage split without featureselection reaches 89.7810% at maximum; 2) Model 2: the accuracy of percentage split with expert feature-selection also reaches 89.7810% at maximum. This study yields the two management implications: 1) ERP vendors can find out hidden potential customers by the proposal models; 2) expert feature-selection of given data is an effective technique used to increase the purpose of classification quality.