Thesis Defense - Beach

Social data is particularly interesting to anonymity research due to its personal nature and recent increase in occurrence and usage. Despite the personal nature and prevalence of social networks and their data, little research has been devoted to their anonymization. This thesis presents work by the author to define and apply anonymity metrics to socio-digital systems.

This thesis discusses: (1) new anonymity definitions that take into account the highly inter-related and often leaked nature of social network data, (2) algorithms to measure these anonymity metrics efficiently, (3) anonymization algorithms which balance information gain with anonymity, (4) a general data model connecting anonymity, information metrics, and knowledge models to make anonymity practical for real world usage and (5) a prototype system which anonymizes Facebook data released over an API to support anonymous socio-digital systems and applications.