BankFair: Balancing Accuracy and Fairness under Varying User Traffic in Recommender System

Driven by sustainability and economic considerations, two-sided recommendation platforms are required to satisfy the needs of both users and providers. Previous studies often indicate that the two sides' needs differ in urgency: providers have relatively long-term exposure requirements, while users desire short-term, accurate services. However, our empirical study reveals that existing methods for balancing fairness and accuracy often fail to ensure both long-term fairness and short-term accuracy under fluctuating user traffic in real applications. Notably, when user traffic is low, user experience tends to decline significantly. Then, we conducted a theoretical analysis confirming that user traffic is a crucial factor in such a trade-off problem. Ensuring accuracy and fairness under variable user traffic remains a challenge. Inspired by the bankruptcy problem in economics, we propose a novel fairness-aware re-ranking approach called BankFair. BankFair intuitively uses the Talmud rule to leverage periods of high user traffic to compensate for periods of low traffic, ensuring consistent user service while maintaining long-term fairness. BankFair is composed of two modules: (1) utilizing the Talmud rule to determine the necessary degree of fairness across varying user traffic periods, and (2) implementing an online re-ranking algorithm based on the fairness degree established by the Talmud rule. Experiments on one publicly available and one real industrial dataset demonstrate that BankFair outperforms all baselines in terms of both accuracy and provider fairness.

A Taxation Perspective for Fair Re-ranking

Fair re-ranking aims to redistribute ranking slots among items more equitably to ensure responsibility and ethics. The exploration of redistribution problems has a long history in economics, offering valuable insights for conceptualizing fair re-ranking as a taxation process. Such a formulation provides us with a fresh perspective to re-examine fair re-ranking and inspire the development of new methods. From a taxation perspective, we theoretically demonstrate that most previous fair re-ranking methods can be reformulated as an item-level tax policy. Ideally, a good tax policy should be effective and conveniently controllable to adjust ranking resources. However, both empirical and theoretical analyses indicate that the previous item-level tax policy cannot meet two ideal controllable requirements: (1) continuity, ensuring minor changes in tax rates result in small accuracy and fairness shifts; (2) controllability over accuracy loss, ensuring precise estimation of the accuracy loss under a specific tax rate. To overcome these challenges, we introduce a new fair re-ranking method named Tax-rank, which levies taxes based on the difference in utility between two items. Then, we efficiently optimize such an objective by utilizing the Sinkhorn algorithm in optimal transport. Upon a comprehensive analysis, Our model Tax-rank offers a superior tax policy for fair re-ranking, theoretically demonstrating both continuity and controllability over accuracy loss. Experimental results show that Tax-rank outperforms all state-of-the-art baselines in terms of effectiveness and efficiency on recommendation and advertising tasks.

LTP-MMF: Towards Long-term Provider Max-min Fairness Under Recommendation Feedback Loops

Multi-stakeholder recommender systems involve various roles, such as users, providers. Previous work pointed out that max-min fairness (MMF) is a better metric to support weak providers. However, when considering MMF, the features or parameters of these roles vary over time, how to ensure long-term provider MMF has become a significant challenge. We observed that recommendation feedback loops (named RFL) will influence the provider MMF greatly in the long term. RFL means that recommender system can only receive feedback on exposed items from users and update recommender models incrementally based on this feedback. When utilizing the feedback, the recommender model will regard unexposed item as negative. In this way, tail provider will not get the opportunity to be exposed, and its items will always be considered as negative samples. Such phenomenons will become more and more serious in RFL. To alleviate the problem, this paper proposes an online ranking model named Long-Term Provider Max-min Fairness (named LTP-MMF). Theoretical analysis shows that the long-term regret of LTP-MMF enjoys a sub-linear bound. Experimental results on three public recommendation benchmarks demonstrated that LTP-MMF can outperform the baselines in the long term.