FRS4CPP: A Fair Recommendation Strategy Considering Interests of Users, Providers and Platform

Haiyan Zhao, Ping Zhou, Jian Cao, Nengjun Zhu

Abstract: Currently, the research on the fairness of recommender systems has expanded beyond considering only the interests of users and product providers. However, the stakeholders of recommender systems go beyond just users and product providers; the platform that provides the recommendation service is also an important player whose interests are currently overlooked in the recommendation algorithm. In this study, we address this gap by considering the potential gains of the platform in addition to recommendation quality and fairness among providers. We analyze the theoretical relationships between recommendation quality, fairness among providers, and the platform’s potential gains. Subsequently, we propose a fair recommendation strategy that takes into account the interests of all three parties. Through experiments conducted on a real-world dataset, we demonstrate that our models successfully achieve the desired design goals.