Modeling Identity-Related Properties and Their Privacy Strength

In the last years several attempts to define identity-related properties such as identifiability, pseudonymity and anonymity have been made to analyze the privacy offered by information systems and protocols. However, these definitions are generally incomparable, making it difficult to generalize the results of their analysis. In this paper, we propose a novel framework for formalizing and comparing identity-related properties. The framework employs the notions of detectability, associability and provability to assess the knowledge of an adversary. We show how these notions can be used to specify well-known identity-related properties and classify them with respect to their logical relations and privacy strength. We also demonstrate that the proposed framework is able to capture and compare several existing definitions of identity-related properties.