ABSTRACT
Market basket analysis (MBA) discovers association patterns in retail transaction data. Marketing applications, such as the product recommendation and the bundling product for promotion, use the discovered patterns to customize strategies. This type of application, referred as to MBA-based applications in this chapter, can have very different formats. For example, the process of product recommendation takes the format of query with condition, which has certain given products and searches for products associated with these given products. But in the process of bundling products, the format is basically query without condition, which has no given product and searches for combinations to increase profits via among-product associations.
