Robust De-anonymization of Large Sparse Datasets

TLDR

The paper introduces a new class of statistical de‑anonymization attacks targeting high‑dimensional micro‑data such as individual preferences and transaction records. The authors apply these attacks to the Netflix Prize dataset of 500,000 anonymous movie ratings. The attacks are robust to data perturbation and background‑knowledge errors, enabling an adversary with limited information to identify individual Netflix records and link them to sensitive attributes such as political preferences.

Abstract

We present a new class of statistical de- anonymization attacks against high-dimensional micro-data, such as individual preferences, recommendations, transaction records and so on. Our techniques are robust to perturbation in the data and tolerate some mistakes in the adversary's background knowledge. We apply our de-anonymization methodology to the Netflix Prize dataset, which contains anonymous movie ratings of 500,000 subscribers of Netflix, the world's largest online movie rental service. We demonstrate that an adversary who knows only a little bit about an individual subscriber can easily identify this subscriber's record in the dataset. Using the Internet Movie Database as the source of background knowledge, we successfully identified the Netflix records of known users, uncovering their apparent political preferences and other potentially sensitive information.