Meta-Learning and Targeted Differential Privacy to Improve the Accuracy-Privacy Trade-off in Recommendations
Peter Müllner, Dominik Kowald, Markus Schedl, Elisabeth Lex
TLDR
This paper improves recommendation accuracy and privacy by combining targeted differential privacy on sensitive data with meta-learning for noise robustness.
Key contributions
- Introduces "targeted DP" to apply privacy only to stereotypical, sensitive user data.
- Uses meta-learning to enhance model robustness against differential privacy noise.
- Significantly improves the accuracy-privacy trade-off in recommender systems.
- Outperforms uniform and full DP baselines in accuracy and empirical privacy risk.
Why it matters
Privacy-preserving recommender systems struggle to balance accuracy with differential privacy. This paper offers a novel solution by combining targeted DP on sensitive data with meta-learning, significantly improving the accuracy-privacy trade-off. This approach is vital for developing more effective and private recommendation engines.
Original Abstract
Balancing differential privacy (DP) with recommendation accuracy is a key challenge in privacy-preserving recommender systems, since DP-noise degrades accuracy. We address this trade-off at both the data and model levels. At the data level, we apply DP only to the most stereotypical user data likely to reveal sensitive attributes, such as gender or age, to reduce unnecessary perturbation; we refer to this as targeted DP. At the model level, we use meta-learning to improve robustness to remaining DP-noise. This achieves a better trade-off between accuracy and privacy than standard approaches: Meta-learning improves accuracy and targeted DP leads to lower empirical privacy risk compared to uniformly applied DP and full DP baselines. Overall, our findings show that selectively applying DP at the data level together with meta-learning at the model level can effectively balance recommendation accuracy and user privacy.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.