A survey of state-of-the-art in anonymity metrics

Abstract

Anonymization enables organizations to protect their data and systems from a diverse set of attacks and preserve privacy; however, in the area of anonymized network data, few, if any, are able to precisely quantify how anonymized their information is for any particular dataset. Indeed, recent research indicates that many anonymization techniques leak some information. An ability to confidently measure this information leakage and any changes in anonymity levels plays a crucial role in facilitating the free-flow of cross-organizational network data sharing and promoting wider adoption of anonyimzation techniques. Fortunately, multiple methods of analyzing anonymity exist. Typical approaches use simple quantifications and probabilistic models; however, to the best of our knowledge, only one network data anonymization metric has been proposed. More importantly, no one-stop-shop paper exists that comprehensively surveys this area for other candidate measures; therefore, this paper explores the state-of-the-art of anonymity metrics. The objective is to provide a macro-level view of the systematic analysis of anonymity preservation, degradation, or elimination for data anonymization as well as network communciations anonymization.