Relational Association Rule Mining in Market Basket using the RoloDex Model with pTree

摘要

In this paper we are concerned with finding how the RoloDex Model can be used to find relational association rules between different entities in Market Basket research using pTrees. The significance of Association rules is measured via support and confidence and they are used to identify the rules in particular transactions. In this paper we will however try to extrapolate that notion into extending it to multiple entities and multi relations using the RoloDex model. RoloDex model is fairly a new concept introduced in this paper. We structure the paper by initially providing some background information on using the notion of Support and Confidence in Market Basket Analysis and then introducing the concept of RoloDex Model and pTrees and finally how the RoloDex model can be used in Market Basket research with pTrees to find multiple relationships between multiple entities.