Scalable and Provably Fair Exposure Control for Large-Scale Recommender Systems (2402.14369v1)
Abstract: Typical recommendation and ranking methods aim to optimize the satisfaction of users, but they are often oblivious to their impact on the items (e.g., products, jobs, news, video) and their providers. However, there has been a growing understanding that the latter is crucial to consider for a wide range of applications, since it determines the utility of those being recommended. Prior approaches to fairness-aware recommendation optimize a regularized objective to balance user satisfaction and item fairness based on some notion such as exposure fairness. These existing methods have been shown to be effective in controlling fairness, however, most of them are computationally inefficient, limiting their applications to only unrealistically small-scale situations. This indeed implies that the literature does not yet provide a solution to enable a flexible control of exposure in the industry-scale recommender systems where millions of users and items exist. To enable a computationally efficient exposure control even for such large-scale systems, this work develops a scalable, fast, and fair method called \emph{\textbf{ex}posure-aware \textbf{ADMM} (\textbf{exADMM})}. exADMM is based on implicit alternating least squares (iALS), a conventional scalable algorithm for collaborative filtering, but optimizes a regularized objective to achieve a flexible control of accuracy-fairness tradeoff. A particular technical challenge in developing exADMM is the fact that the fairness regularizer destroys the separability of optimization subproblems for users and items, which is an essential property to ensure the scalability of iALS. Therefore, we develop a set of optimization tools to enable yet scalable fairness control with provable convergence guarantees as a basis of our algorithm.
- Multistakeholder recommendation: Survey and research directions. In UMUAI.
- Controlling popularity bias in learning-to-rank recommendation. In RecSys.
- Managing popularity bias in recommender systems with personalized re-ranking. In International FLAIRS Conference.
- Anthony B Atkinson et al. 1970. On the measurement of inequality. In Journal of Economic Theory.
- A generic coordinate descent framework for learning from implicit feedback. In WWW.
- The million song dataset. In ISMIR.
- Equity of attention: Amortizing individual fairness in rankings. In SIGIR.
- Distributed optimization and statistical learning via the alternating direction method of multipliers. In Foundations and Trends® in Machine learning.
- Goodbye pareto principle, hello long tail: The effect of search costs on the concentration of product sales. In Management Science.
- Robin Burke. 2017. Multisided fairness for recommendation. In arXiv preprint arXiv:1707.00093.
- Balanced neighborhoods for multi-sided fairness in recommendation. In FAccT.
- The direct extension of ADMM for multi-block convex minimization problems is not necessarily convergent. In Mathematical Programming.
- Revisiting negative sampling vs. non-sampling in implicit recommendation. In TOIS.
- Lorslim: Low rank sparse linear methods for top-n recommendations. In ICDM.
- Parallel multi-block ADMM with O (1/k) convergence. In Journal of Scientific Computing.
- Two-sided fairness in rankings via Lorenz dominance. In NIPS.
- Virginie Do and Nicolas Usunier. 2022. Optimizing generalized Gini indices for fairness in rankings. In SIGIR.
- John Duchi and Yoram Singer. 2009. Efficient online and batch learning using forward backward splitting. In JMLR.
- Daniel Fleder and Kartik Hosanagar. 2009. Blockbuster culture’s next rise or fall: The impact of recommender systems on sales diversity. In Management Science.
- Marguerite Frank and Philip Wolfe. 1956. An algorithm for quadratic programming. In Naval Research Logistics Quarterly.
- F Maxwell Harper and Joseph A Konstan. 2015. The movielens datasets: History and context. In TiiS.
- Neural collaborative filtering. In WWW.
- Exposure diversity as a design principle for recommender systems. In Information, Communication & Society.
- Collaborative filtering for implicit feedback datasets. In ICDM.
- Coupled graphs and tensor factorization for recommender systems and community detection. In TKDE.
- Martin Jaggi. 2013. Revisiting Frank-Wolfe: Projection-free sparse convex optimization. In ICML.
- Efficiency Improvement of Neutrality-Enhanced Recommendation. In Human Decision Making in Recommender Systems (Decisions@ RecSys’ 13).
- Fairness-aware learning through regularization approach. In ICDM.
- Variational autoencoders for collaborative filtering. In WWW.
- Linearized ADMM for nonconvex nonsmooth optimization with convergence analysis. In IEEE Access.
- Paolo Massa and Paolo Avesani. 2007. Trust-aware recommender systems. In RecSys.
- Fairness for Robust Learning to Rank. In arXiv preprint arXiv:2112.06288.
- Controlling fairness and bias in dynamic learning-to-rank. In SIGIR.
- Harrie Oosterhuis. 2021. Computationally efficient optimization of plackett-luce ranking models for relevance and fairness. In SIGIR.
- Yoon-Joo Park and Alexander Tuzhilin. 2008. The long tail of recommender systems and how to leverage it. In RecSys.
- Fairrec: Two-sided fairness for personalized recommendations in two-sided platforms. In WWW.
- Robin L Plackett. 1975. The analysis of permutations. In Journal of the Royal Statistical Society: Series C (Applied Statistics).
- Neural collaborative filtering vs. matrix factorization revisited. In RecSys.
- Revisiting the Performance of iALS on Item Recommendation Benchmarks. In arXiv preprint arXiv:2110.14037.
- Bursting your (filter) bubble: strategies for promoting diverse exposure. In CSCW.
- R Tyrrell Rockafellar. 1976. Monotone operators and the proximal point algorithm. In SIAM Journal on Control and Optimization.
- Yuta Saito. 2020. Unbiased pairwise learning from biased implicit feedback. In Proceedings of the 2020 ACM SIGIR on International Conference on Theory of Information Retrieval. 5–12.
- Yuta Saito and Thorsten Joachims. 2022. Fair Ranking as Fair Division: Impact-Based Individual Fairness in Ranking. In KDD.
- Unbiased recommender learning from missing-not-at-random implicit feedback. In Proceedings of the 13th International Conference on Web Search and Data Mining. 501–509.
- Exploiting query reformulations for web search result diversification. In WWW.
- Jack Sherman and Winifred J Morrison. 1950. Adjustment of an inverse matrix corresponding to a change in one element of a given matrix. In The Annals of Mathematical Statistics.
- Ashudeep Singh and Thorsten Joachims. 2018. Fairness of exposure in rankings. In KDD.
- Ashudeep Singh and Thorsten Joachims. 2019. Policy learning for fairness in ranking. In NIPS.
- Constrained tensor factorization with accelerated AO-ADMM. In ICPP.
- Harald Steck. 2018. Calibrated recommendations. In RecSys.
- Admm slim: Sparse recommendations for many users. In WSDM.
- Harald Steck and Dawen Liang. 2021. Negative Interactions for Improved Collaborative Filtering: Don’t go Deeper, go Higher. In RecSys.
- Fast online ranking with fairness of exposure. In FAccT.
- Parallel direction method of multipliers. In NIPS.
- Max A Woodbury. 1950. Inverting modified matrices. Statistical Research Group.
- A Multi-objective Optimization Framework for Multi-stakeholder Fairness-aware Recommendation. In TOIS.
- TFROM: A two-sided fairness-aware recommendation model for both customers and providers. In SIGIR.
- Policy-gradient training of fair and unbiased ranking functions. In SIGIR.
- Sirui Yao and Bert Huang. 2017. Beyond parity: Fairness objectives for collaborative filtering. In NIPS.
- Parallel matrix factorization for recommender systems. In Knowledge and Information Systems.
- Distributed stochastic ADMM for matrix factorization. In CIKM.
- Meike Zehlike and Carlos Castillo. 2020. Reducing disparate exposure in ranking: A learning to rank approach. In WWW.
- Large-scale parallel collaborative filtering for the netflix prize. In AAIM.