Tuning Pseudonymization Parameters in a Privacy by Design Approach for Secure Information Discovery Between Federated Organizations

Secure and privacy-preserving information discovery between independent organizations is a major challenge in a variety of contexts, such as public authorities (e.g., public security or health authorities) or business cooperation of private organizations. Isolated databases, security and privacy regulations, as well as a lack of trust often hinder an efficient and secure information discovery and information exchange between these entities. To overcome these obstacles, a generic semi-automated hit/no-hit-system for secure information discovery between federated organizations has been developed in a joint research project that is presented in this publication. The distributed system uses a pseudonymization approach allowing for similarity searches between pseudonyms. It has been tested in a pilot phase of the European Police Records Index System (EPRIS) which will become the basis for the new regulation for automated data exchange for police cooperation in the EU. This paper presents the underlying considerations of choosing and tuning the pseudonymization parameters.