Membership Privacy: A Unifying Framework For Privacy Definitions

Oct 30, 2013

Download: MP4 Video Size: 146.2MB  
Watch on YouTube

Abstract

Data collected by organizations and agencies are a key resource
in today's information age. The use of sophisticated data mining techniques
makes it possible to extract relevant knowledge that can then be used for a
variety of purposes, such as research, developing innovative technologies
and services, intelligence and counter-terrorism operations, and providing
inputs to public policy making. However the disclosure of those data poses
serious threats to individual privacy. In this talk, we present a novel
privacy framework that we call Membership Privacy, which prevents the
adversary from significantly increasing its ability to conclude that an
entity is in the input dataset. Membership privacy is parameterized by a
family of distributions that captures the adversary's prior knowledge. The
power and flexibility of the proposed framework lies in the ability to
choose different distribution families to instantiate membership privacy.
Many privacy notions in the literature are equivalent to membership privacy
with interesting distribution families, including differential privacy,
differential identifiability, and differential privacy under sampling. The
framework also provides a principled approach to developing new privacy
notions under which better utility can be achieved than what is possible
under differential privacy. This is joint work with Wahbeh Qardaji, Dong
Su, Yi Wu, and Weining Yang.