Privacy-Preserving Data Publication for Static and Streaming Data

Abstract

The thesis concentrates on the anonymization of static and streaming data. In the static settings, we first present SABRE, a Sensitive Attribute Bucketization and REdistribution framework for t-closenes. Then, we propose beta-likeness, a privacy model beyond t-closeness, which postulates that an adversary's confidence on the likelihood of a certain sensitive value should not increase, in relative difference terms, by more than a predefined threshold. In the context of data streams, we present CASTLE (Continuously Anonymizing STreaming data via adaptive cLustEring), a cluster-based scheme that continuously anonymizes data streams and, at the same time, ensures the freshness of the anonymized data by satisfying specified delay constraints. To better protect the privacy of streaming data, we have also customized t-closeness and applied it to data streams. The extensive experimental results show that our solutions achieve information quality superior to existing ones, and can be faster as well.