Hybrid Product Recommender System for Apparel Retailing Customers

In apparel retailing stores, new shopping experience provided to customers such as automatic product recommender could improve the brand loyalty. In this paper, we propose a hybrid recommender system consisting of both content-based collaborative filtering and collaborative filtering so as to provide buying recommender for the VIP customers of apparel retailing stores. Firstly, the VIP customer who holds a RFID-circuit-embedded VIP card is automatically identified via the customer identifying system. After that, the hybrid recommender system would generate personalized recommender list based on content-based filtering and collaborative filtering strategy. The contents on the list are from the following three different resources: the products that are similar to or in the same style with what the objective customer have selected before in this store, the products that other VIP members who has the same tastes with the objective customer have chosen recently and the products that are similar to or in the same style with those VIP members have bought recently. After briefly review the related work of recommender system, this paper illustrates the structure of the system. It then analyzes the algorithm of the content-based filtering and collaborative filtering strategy in the system. Experimental result shows that the hybrid recommender system could implement the customer tastes analysis and the product recommender in apparel retailing store.