Privacy-preserving Data Publishing: RAPPOR and Differential Privacy

This lecture covers the RAPPOR mechanism for privacy-preserving data publishing, focusing on Differential Privacy. It explains the steps involved in RAPPOR, such as mapping URLs to Bloom filters, Permanent and Instant Randomized Response. It also discusses Apple's implementation of Differential Privacy, k-means clustering, and making k-means differentially private. The lecture concludes with a discussion on differentially private synthetic data and the challenges of implementing differential privacy in practice.