Data privacy preservation at digital information age

点击翻译按钮获取中文摘要

The goal of this proposal is to investigate and develop techniques for privacy preserving data publishing (PPDP), i.e., allowing publication of data for the benefit of research while preserving individual privacy. The basic form of the problem we will study is as follows. A publisher wants to publish some person-specific data (e.g, a medical database) to some (possibly unknown) recipient for research purposes. The data contains private information that is essential to research. The recipient may be "curious" trying to re-identify the individual corresponding to a data record, possibly with the help of some prior knowledge. The recipient with this behavior is called "adversary". To prevent re-identification of data subjects, data must be anonymized so that data subjects cannot be associated with their records with a high probability. The key question is how to anonymize the data so that individual privacy is respected and the anonymized data remains practically useful to research.