Forecasting Personal Shopping Behavior

Abstract

Data mining (DM) techniques make efforts to discovery knowledge from data. Aiming to finding patterns, association rule (AR) computing algorithms seem to be one to be adopted on variety applications. To be originally claimed for best analyzing customer shopping goods in baskets, Apriori, the first AR algorithm, has been discussed and modified the most by researchers. This study adopts Apriori algorithm to forecast individual customer shopping behavior. This study finds that customer shopping behaviors can be comprehended better in a long run. With Apriori mining and the examining principles proposed by this study, customer purchase behaviors of no matter constant purchase, stopping purchasing habitual goods, and starting to purchase goods that never bought before is able to be recognized. However, impulse purchase, including purchase for holidays, is unable to be discovered.