Anonymization of statistical data

In the modern digital society, personal information about individuals can be collected, stored, shared and disseminated much more easily and freely. Such data can be released in “macrodata” form, reporting aggregated information, or in “microdata” form, reporting specific information on individual respondents. To ensure proper privacy of individuals as well of public and private organizations, it is then important to protect possible sensitive information in the original dataset from either direct or indirect disclosure. In this paper, we characterize macrodata and microdata releases and then focus on microdata protection. We provide a characterization of the main microdata protection techniques and describe recent solutions for protecting microdata against identity and attribute disclosure, discussing some open issues that need to be investigated.