2009 DifferentiallyPrivateRecommende

• (McSherry et al., 2009) ⇒ Frank McSherry, and Ilya Mironov. (2009). “Differentially Private Recommender Systems: Building Privacy Into the Net.” In: Proceedings of the 15th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD-2009). doi:10.1145/1557019.1557090

Subject Headings:

Notes

• Categories and Subject Descriptors: H.2.8 Database Management: Data Mining.
• General Terms: Algorithms, Security, Theory

Cited By

• http://scholar.google.com/scholar?q=%22Differentially+private+recommender+systems%3A+building+privacy+into+the+net%22+2009
• http://portal.acm.org/citation.cfm?doid=1557019.1557090&preflayout=flat#citedby

Quotes

Author Keywords

Differential Privacy, Netflix, Recommender Systems

Abstract

We consider the problem of producing recommendations from collective user behavior while simultaneously providing guarantees of privacy for these users. Specifically, we consider the Netflix Prize data set, and its leading algorithms, adapted to the framework of differential privacy.
Unlike prior privacy work concerned with cryptographically securing the computation of recommendations, differential privacy constrains a computation in a way that precludes any inference about the underlying records from its output. Such algorithms necessarily introduce uncertainty i.e., noise to computations, trading accuracy for privacy.
We find that several of the leading approaches in the Netflix Prize competition can be adapted to provide differential privacy, without significantly degrading their accuracy. To adapt these algorithms, we explicitly factor them into two parts, an aggregation/learning phase that can be performed with differential privacy guarantees, and an individual recommendation phase that uses the learned correlations and an individual's data to provide personalized recommendations. The adaptations are non-trivial, and involve both careful analysis of the per-record sensitivity of the algorithms to calibrate noise, as well as new post-processing steps to mitigate the impact of this noise.
We measure the empirical trade-off between accuracy and privacy in these adaptations, and find that we can provide non-trivial formal privacy guarantees while still outperforming the Cinematch baseline Netflix provides.

References

,

	Author	volume	Date Value	title	type	journal	titleUrl	doi	note	year
2009 DifferentiallyPrivateRecommende	Frank McSherry Ilya Mironov			Differentially Private Recommender Systems: Building Privacy Into the Net		KDD-2009 Proceedings		10.1145/1557019.1557090		2009