This strategic project will address in particular the issue of privacy by design in big data infrastructures, with special emphasis on purpose limitation and anonymization. Together with our scientific and our industrial partners we will adapt anonymization methodologies and purpose relaxation concepts as weil as approaches for transparency enforcement towards data subjects. Objective 1: - Maximization of automatized transparency towards data subjects and of purpose compatibility: Embedded in a completely transparent, privacy engineered big data framework, data subjects shall be informed about the current usage of their personal data at any time. Furthermore, data subjects shall also be informed about potential further usages of their data, so that they can freely decide at any time which of their personal data should be processed for which purpose. This objective goes hand in hand with the design of a flexible access control and privacy rights management which ensures that data processing is limited to legitimate purposes by authenticated persons only. Additionally, criteria and guidelines shall be developed that allow for easy verification if the processing of data for new purposes -which might be often the case in Big Data applications where data from different sources are analysed for previously unknown purposes -is compatible with the original purpose. Objective 2: -Adaption and optimization of anonymization concepts for big data infrastructures: Anonymization is an important method in order to comply with the principle of data minimisation. Appropriately anonymized data may be used freely without having to consider data protection law. Based on well-known concepts, such as k-anonymity, t-closeness and 1- diversity, we aim at applying, adapting and optimizing these methods in the context of privacy engineered data management.
|Effective start/end date||1/06/19 → 31/12/19|
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.