TK researcher Aidmar Wainakh presenting Paper at RecSys 2019

In Copenhagen, Aidmar presentsed his collaborative work with Tim, Jörg, and Max on the establishment of privacy-preserving recommendations using a combination of user-sampling and data-retention to the research community at ACM's annual conference on Recommender Systems. With their approach, the complexity of computation and computation is reduced, as computation-heavy cryptographic primitives are avoided and the set of contributing users is reduced. While the amount of data is minimized, the precision and recall of derived association rules is maintained at a high level while the communication cost could be reduced by 50%.

For more details, please contact Aidmar Wainakh (wainakh@privacy-trust.tu-…) or refer to the full paper (https://dl.acm.org/citation.cfm?id=3347013)