Learning and Sharing under Privacy constraints (DataLEASH)

About the project
DataLEASH will develop and test methods for more open data. In practice, it will consist of risk analysis for privacy protection (key indicators, methodology, legal requirements) and privacy configured learning systems (mechanisms, considerations of integrity and user value). It will support public organizations who are required to have open data and therefore need managing methods that are fast, reliable and uncomplicated. Based on these methods they are able to make well-informed decisions on how and if data should be shared (limiting access and various security settings) and choose what form of data conversion that should be aligned with a certain level of privacy and use.

Background
Privacy and managing sensitive data is a major challenge and growing with the increasing collections of data and advanced systems based on machine learning for public services. Legal restrictions e.g GDPR has improved privacy for managing personal data but there are still no guidelines for how to restrict data leakage and this has implications for how data can be shared and used.

Cross-disciplinary collaboration
DataLEASH brings together researchers from the School of Electrical Engineering and Computer Science (EECS, KTH), from the Department of Law at Stockholm University and from the Decisions, Network, and Analytics lab at RISE.