Customized Bundle Recommendation by Association Rules of Product Categories for Online Supermarkets

A customized bundle is a list of products recommended to consumers among them they can choose his/her favorite products according to his/her preference. It is an efficient way to not only simplify the customer's shopping process, but also reduce the order fulfillment cost for the online supermarkets. A customized bundle recommendation method is proposed for online supermarket in this research. It is realized by combinational using association rule mining, customer segmentation and recommendation techniques. The association rules of product category are used to avoid mass unnecessary association rules of product. The product lists recommended within each category are generated by product ranking on each customer segmentation. Numerical experiments are conducted to verify the effectiveness of the proposed method. The method can be easily extended.