Fairness-aware Differentially Private Collaborative Filtering

Abstract

Recently, there has been an increasing adoption of differential privacy guided algorithms for privacy-preserving machine learning tasks. However, the use of such algorithms comes with trade-offs in terms of algorithmic fairness, which has been widely acknowledged. Specifically, we have empirically observed that the classical collaborative filtering method, trained by differentially private stochastic gradient descent (DP-SGD), results in a disparate impact on user groups with respect to different user engagement levels. This, in turn, causes the original unfair model to become even more biased against inactive users. To address the above issues, we propose \textbf{DP-Fair}, a two-stage framework for collaborative filtering based algorithms. Specifically, it combines differential privacy mechanisms with fairness constraints to protect user privacy while ensuring fair recommendations. The experimental results, based on Amazon datasets, and user history logs collected from Etsy, one of the largest e-commerce platforms, demonstrate that our proposed method exhibits superior performance in terms of both overall accuracy and user group fairness on both shallow and deep recommendation models compared to vanilla DP-SGD.

References in corpus (13)

• Deep Learning with Differential Privacy
• BPR: Bayesian Personalized Ranking from Implicit Feedback
• Ups and Downs: Modeling the Visual Evolution of Fashion Trends with One-Class Collaborative Filtering
• Learning Differentially Private Recurrent Language Models
• User-oriented Fairness in Recommendation
• Towards Long-term Fairness in Recommendation
• Beyond Parity: Fairness Objectives for Collaborative Filtering
• Understanding Echo Chambers in E-commerce Recommender Systems
• A General Approach to Adding Differential Privacy to Iterative Training Procedures
• Differential Privacy Has Disparate Impact on Model Accuracy
• A Survey on Trustworthy Recommender Systems
• Variation Control and Evaluation for Generative SlateRecommendations
• Differentially Private SGDA for Minimax Problems