Privacy-preserving data publishing

Objectifs

"Personal data is the new oil of the Internet and the new currency of the digital world", declared M. Kouneva, European Commissioner for Consumer Protection in March 2009. The value of massive personal data analysis for industry, science and society in general is widely recognized today. However, their personal and potentially sensitive nature is a major obstacle to their large-scale sharing. The aim of privacy-preserving data publication models and algorithms is precisely to offer strong guarantees of privacy while allowing quality sharing for analysis purposes. The task is far from trivial, as several re-identification scandals have shown. The aim of this course is to introduce students to the main paradigms and techniques of privacy-friendly data publication.

Particular emphasis will be placed on one of today’s most prominent models: differential privacy.

Plan

Presentations by various external speakers. These presentations will showcase individual data processing and publication issues and solutions in their industrial context.

Prérequis

Basic knowledge of data management, algorithms, probability and statistics & Basic skills in a programming language such as Java, Python or R.