How Much High Level Utility in Anonymized Social Networks?
Primary Investigator:
Feng Li
Tianchong Gao; Feng Li
Abstract
The traditional differential privacy anonymization techniques are proved to maintain some utility metrics. However, the performance of the anonymized graph is doubtful in read application. A new high level utility, persistent homology, is presented to evaluate these techniques. And the project hopes to give insights of designing new anonymization techniques.