A1077
Title: Integrating privacy enhancements with dynamic community detection
Authors: Liyan Xie - University of Minnesota (United States) [presenting]
Abstract: In the evolving landscape of online communities, safeguarding user privacy while accurately detecting dynamic changes presents a critical challenge. The private online community detection problem is studied by integrating edge differential privacy (DP) within the framework of a censored block model (CBM). The fundamental tradeoffs between the privacy budget, detection performance, and exact community recovery of community labels are explored. The proposed algorithm can identify changes in the community structure while maintaining user privacy. A new information-theoretic lower bound is also established on the delay in detecting community changes privately.
