The problem of statistical disclosure control - revealing accurate statistics about a population while preserving the privacy of individuals - has a venerable history. An extensive literature spans multiple disciplines: statistics, theoretical computer science, security, and databases.
This talks describes a large body of work revisiting the problem from a fresh perspective. We present a powerful technique for achieving a mathematically rigorous and comprehensive notion of privacy while simultaneously preserving utility of the data, together with impossibility results that guided its development.
Reception to follow in 301 Coordinated Science Laboratory.
Cynthia Dwork has made fundamental contributions to complexity theory, distributed computing, and cryptography. Her current focus is the development of a mathematically rigorous framework for the privacy-preserving analysis of data. Together with several collaborators, she has articulated new and powerful privacy definitions, and applied them to data analysis problems ranging from OLAP reporting to machine learning. A Principal Researcher at Microsoft, Dwork has previously held positions at the IBM Almaden Research Center and the Compaq Systems Research Center. Within the past year, her work has received the Best Paper Award at the World Wide Web Conference and the Edsger W. Dijkstra Prize in Distributed Computing.