Privacy in Data Publishing

In modern digital society, personal information about individuals can be easily collected, shared, and disseminated. These data collections often contain sensitive information, which should not be released in association with respondents' identities. Removing explicit identifiers before data release does not offer any guarantee of anonymity, since de-identified datasets usually contain information that can be exploited for linking the released data with publicly available collections that include respondents' identities. To overcome these problems, new proposals have been developed to guarantee privacy in data release. In this chapter, we analyze the risk of disclosure caused by public or semi-public microdata release and we illustrate the main approaches focusing on protection against unintended disclosure. We conclude with a discussion on some open issues that need further investigation.