Proposing novel algorithm for basket analysis using consistent association rules

Abstract

Market Basket Analysis is a technique based on a theory that if someone buys one group of items many times together then, he/she is more or less likely to purchase those items together in the future purchases rather than buy those items in isolation. For example, if one set of customers keep buying beer, coke, and diapers together, it is highly likely that when they enter the store the next time, if they purchase beer & coke, they would also buy diapers. The set of items or products bought together is called an itemset, and market Basket Analysis aims to find relationships amongst those purchases.