A Recommender System for E-Commerce with Strategy-Oriented Modules based on Insufficient Information

Abstract

Electronic Commerce (EC) has become an important support for business and is regarded as an efficient system that connects suppliers with online users. Among the applications of EC, a Recommender System (RS) is undoubtedly a popular approach for promoting the products actively to the users. Even if many approaches have been proposed, a comprehensive module comprising of essential sub-modules of input profiles, a recommendation scheme, and an output interface of recommendations in the RS is still lacking. Besides, many approaches are confronted with the cold-start problem, which can be attributed to the problem of sparse user-item matrices. In addition, the fundamental issue of profit consideration for an EC company is not addressed in general terms. Therefore, this thesis aims to construct an RS with a strategy-oriented operation module regarding the above aspects; and with this module, three sub-modules of input, association prediction and output are proposed along with three tools of the Expectation-Robust Principal Component Analysis (E-RPCA), the Clique-Effects Collaborative Filtering (CECF), and the Strategy Analysis Model (SAM). The proposed RS module has been implemented by several experiments and the case studies of 3C retailers in Taiwan; promising results were obtained to approve the contributions of the proposed module.