CONDITIONAL ASSOCIATION RULES FOR MINING THE FEATURES OF CUSTOMER PURCHASES

Abstract

Association analysis is analysis technique for transaction data that etracts the patterns of what items one customer purchases together. By limiting the analysis to specific group of customers, we may extract a rule not among the association rules for the whole. In this study, we used this method, which we refer to as “Conditioning Association Rules”, to analyze trends in purchases of preferred customers characteristics extracted as Conditioning Association Rules for each group. In addition, we analyzed generational and gender differences in purchase trends.