Download now Free registration required
To seek better prediction techniques, data owners of recommender systems such as Netflix sometimes make their customers' reviews available to the public, which raises serious privacy concerns. With only a small amount of knowledge about individuals in a recommender system, an adversary may be able to re-identify users and consequently determine their item ratings. In this paper, the authors present a robust and efficient anonymization algorithm for publishing recommendation datasets, Predictive Anonymization, that gives desired privacy guarantees without significantly affecting prediction accuracy.
- Format: PDF
- Size: 38.1 KB