Differential privacy: Science provides researchers and census-takers a better way to protect personal data

Can the privacy of individual data truly be protected? When it comes to most of the personal information collected from Internet users, the answer at the moment may be no. And in the world of research data — health and social science studies — several proposed ways of protecting personal data have left individuals vulnerable to “re-identification” within data sets. However, computer scientist Cynthia Dwork and her colleagues have solved these problems with a new approach, called differential privacy, that is now being used in many research studies and will be used by the U.S. Census Bureau in the upcoming decennial census. In the past the bureau has relied on “security by obscurity,” using various secret methods to make identification of individuals difficult. Dwork will explain how differential privacy works, why it is a good idea, and why it will allow the Census Bureau to take a radically different approach, revealing its privacy algorithm to all.

Social media hashtag: #DifferentialPrivacy