PRIVACY & CONFIDENTIALITY, DE-ANONYMIZATION

The Objectives

Work tasks

In WP5 we develop solutions for different kinds of scenarios where data owners can securely compute on the combined datasets preserving its privacy and confidentially. In the course of this computation no party should be able to learn any information about the data of the other party except the result of their joint computation. This should lead to specialized protocols, e.g. for business partners to capitalize on their data sets by performing analysis on the combined data without sharing any confidential content or for manufacturers and their customers to interactively negotiate product orders while preserving the confidentiality of the involved business data in such negotiations. In particular, this will facilitate companies to safely engage in big data marketplaces and compute on personal data while ideally maintaining GDPR compliance.
In the last step, the internal verification is important. We have to make sure that data exposed in the project are not reasonably likely to be re-identified. This verification is done by applying a battery of statistical de-anonymization methods and will be considered as successful only when all applied de-anonymization algorithms fail.