Abstract
Collaborative filtering (CF) techniques learn user and item embeddings from user-item interaction behaviors, and are commonly used in recommendation systems to help users find potentially desirable items. Most CF models optimize recommendation accuracy; however, they may lead to unwanted biases for particular demographic groups. Thus, we focus on learning fair representations of CF-based recommendations. We formulate this problem as an optimization task with two competing goals: embedding representations better meet accuracy requirements of recommendations, and simultaneously obfuscate information hidden in the embedding space, which is related to the users’ sensitive attributes for fairness. Here, the intuitive idea is to use fair representation learning from machine learning to train a classifier with a sensitive attribute predictor from the user side to satisfy the fairness goal. However, such fair machine learning models assume entity independence, which differs greatly from CF because users and items are correlated collaboratively via user-item behaviors. Therefore, sensitive user information can be exposed from the users’ preferred items. Consequently, defining only fairness constraints on users cannot achieve fairness in recommendation systems. In this paper, we propose FairCF framework for fairness-aware collaborative filtering. In particular, we first define fairness constraints in a fair embedding space, where both a user classifier and an item classifier are employed to fit the fairness constraints. We then design an item classifier without item sensitive labels. The proposed framework can be trained in an end-to-end manner under most embedding based CF models. Extensive experiments conducted on three datasets (MovieLens-100K, MovieLens-1M, and Lastfm-360K) clearly demonstrate the superiority of the proposed FairCF framework relative to various fairness metrics (i.e., performance of newly-trained classifiers) than other state-of-the-art fairness-aware CF models with less than 4% accuracy reduction.
Similar content being viewed by others
References
van den Oord A, Dieleman S, Schrauwen B. Deep content-based music recommendation. In: Proceedings of Neural Information Processing Systems, 2013. 2643–2651
Wu J W, Shen L W, Guo W N, et al. Code recommendation for android development: how does it work and what can be improved? Sci China Inf Sci, 2017, 60: 092111
Ying R, He R N, Chen K F, et al. Graph convolutional neural networks for web-scale recommender systems. In: Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2018. 974–983
Chen H H, Jin H, Cui X L. Hybrid followee recommendation in microblogging systems. Sci China Inf Sci, 2017, 60: 012102
Mnih A, Salakhutdinov R R. Probabilistic matrix factorization. In: Proceedings of Neural Information Processing Systems, 2008. 1257–1264
Rendle S, Freudenthaler C, Gantner Z, et al. BPR: Bayesian personalized ranking from implicit feedback. In: Proceedings of the 25th Conference on Uncertainty in Artificial Intelligence, 2009. 452–461
He X N, Liao L Z, Zhang H W, et al. Neural collaborative filtering. In: Proceedings of the 26th International Conference on World Wide Web, 2017. 173–182
Lei C, Le W, Richang H, et al. Revisiting graph based collaborative filtering: a linear residual graph convolutional network approach. In: Proceedings of AAAI Conference on Artificial Intelligence, 2020. 27–34
Wu L, Yang Y H, Zhang K, et al. Joint item recommendation and attribute inference: an adaptive graph convolutional network approach. In: Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval, 2020. 679–688
Koren Y, Bell R, Volinsky C. Matrix factorization techniques for recommender systems. Computer, 2009, 42: 30–37
Lahoti P, Gummadi K P, Weikum G. iFair: learning individually fair data representations for algorithmic decision making. In: Proceedings of the 35th International Conference on Data Engineering, 2019. 1334–1345
Wu C H, Wu F Z, Wang X T, et al. Fairness-aware news recommendation with decomposed adversarial learning. In: Proceedings of AAAI Conference on Artificial Intelligence, 2021. 4462–4469
Sweeney L. Discrimination in online ad delivery. SSRN J, 2013, 11: 10–29
Pedreshi D, Ruggieri S, Turini F. Discrimination-aware data mining. In: Proceedings of the 14th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2008. 560–568
Kamiran F, Calders T. Classifying without discriminating. In: Proceedings of the 2nd International Conference on Computer, Control and Communication, 2009. 1–6
Luong B T, Ruggieri S, Turini F. k-NN as an implementation of situation testing for discrimination discovery and prevention. In: Proceedings of the 17th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2011. 502–510
Zhang B H, Lemoine B, Mitchell M. Mitigating unwanted biases with adversarial learning. In: Proceedings of AAAI/ACM Conference on AI, Ethics, and Society, 2018. 335–340
Kamishima T, Akaho S, Sakuma J. Fairness-aware learning through regularization approach. In: Proceedings of the 11th International Conference on Data Mining Workshops, 2011. 643–650
Zemel R, Wu Y, Swersky K, et al. Learning fair representations. In: Proceedings of the 30th International Conference on International Conference on Machine Learning, 2013. 325–333
Edwards H, Storkey A J. Censoring representations with an adversary. In: Proceedings of International Conference on Learning Representations, 2016
Beutel A, Chen J L, Zhao Z, et al. Data decisions and theoretical implications when adversarially learning fair representations. In: Proceedings of Workshop on Fairness, Accountability, and Transparency in Machine Learning, 2017
Madras D, Creager E, Pitassi T, et al. Learning adversarially fair and transferable representations. In: Proceedings of the 35th International Conference on Machine Learning, 2018. 3381–3390
Yao S R, Huang B. Beyond parity: fairness objectives for collaborative filtering. In: Proceedings of the 31st Conference on Neural Information Processing System, 2017. 2921–2930
Bose A J, Hamilton W. Compositional fairness constraints for graph embeddings. In: Proceedings of the 36th International Conference on Machine Learning, 2019. 715–724
Goodfellow I, Pouget-Abadie J, Mirza M, et al. Generative adversarial nets. In: Proceedings of Neural Information Processing Systems, 2014. 2672–2680
Covington P, Adams J, Sargin E. Deep neural networks for youtube recommendations. In: Proceedings of the 10th ACM Conference on Recommender Systems, 2016. 191–198
Paparrizos I, Cambazoglu B B, Gionis A. Machine learned job recommendation. In: Proceedings of the 5th ACM Conference on Recommender systems, 2011. 325–328
Wu L, Ge Y, Liu Q, et al. Modeling the evolution of users’ preferences and social links in social networking services. IEEE Trans Knowl Data Eng, 2017, 29: 1240–1253
Wu L, Chen L, Hong R C, et al. A hierarchical attention model for social contextual image recommendation. IEEE Trans Knowl Data Eng, 2020, 32: 1854–1867
Mehrabi N, Morstatter F, Saxena N, et al. A survey on bias and fairness in machine learning. ACM Comput Sur, 2021, 54: 1–35
Mukherjee D, Yurochkin M, Banerjee M, et al. Two simple ways to learn individual fairness metrics from data. In: Proceedings of the 37th International Conference on Machine Learning, 2020. 7097–7107
Kusner M J, Loftus J, Russell C, et al. Counterfactual fairness. In: Proceedings of Neural Information Processing Systems, 2017. 4066–4076
Hardt M, Price E, Srebro N. Equality of opportunity in supervised learning. In: Proceedings of Neural Information Processing Systems, 2016. 3315–3323
Binns R. Fairness in machine learning: lessons from political philosophy. In: Proceedings of Conference on Fairness, Accountability, and Transparency, 2018. 149–159
Dai E Y, Wang S H. Say no to the discrimination: learning fair graph neural networks with limited sensitive attribute information. In: Proceedings of the 14th ACM International Conference on Web Search and Data Mining, 2021. 680–688
Zhu Z W, Hu X, Caverlee J. Fairness-aware tensor-based recommendation. In: Proceedings of the 27th ACM International Conference on Information and Knowledge Management, 2018. 1153–1162
Beutel A, Chen J L, Doshi T, et al. Fairness in recommendation ranking through pairwise comparisons. In: Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2019. 2212–2220
Islam R, Keya K N, Zeng Z Q, et al. Debiasing career recommendations with neural fair collaborative filtering. In: Proceedings of Web Conference, 2021. 3779–3790
Le W, Lei C, Shao P Y, et al. Learning fair representations for recommendation: a graph-based perspective. In: Proceedings of Web Conference, 2021. 2198–208
Ekstrand M D, Tian M, Azpiazu I M, et al. All the cool kids, how do they fit in?: Popularity and demographic biases in recommender evaluation and effectiveness. In: Proceedings of Conference on Fairness, Accountability and Transparency, 2018. 172–186
Lambrecht A, Tucker C. Algorithmic bias? An empirical study of apparent gender-based discrimination in the display of STEM career ads. Manage Sci, 2019, 65: 2966–2981
Wang Y X, Wu C S, Herranz L, et al. Transferring GANs: generating images from limited data. In: Proceedings of European Conference on Computer Vision, 2018. 220–236
Harper F M, Konstan J A. The movielens datasets: history and context. ACM Trans Interact Intell Syst, 2015, 5: 1–19
Herrada Ò C. Music recommendation and discovery in the long tail. Dissertation for Ph.D. Degree. Barcelona: Universitat Pompeu Fabra, 2009
Acknowledgements
This work was supported in part by National Natural Science Foundation of China (Grant Nos. 61972125, U19A2079, 61725203, 61732008, 62006066) and Fundamental Research Funds for the Central Universities (Grant No. JZ2020HGPA-0114). Le WU greatly thanks the support of Young Elite Scientists Sponsorship Program by CAST and ISZS.
Author information
Authors and Affiliations
Corresponding authors
Rights and permissions
About this article
Cite this article
Shao, P., Wu, L., Chen, L. et al. FairCF: fairness-aware collaborative filtering. Sci. China Inf. Sci. 65, 222102 (2022). https://doi.org/10.1007/s11432-020-3404-y
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1007/s11432-020-3404-y