Group on Recommender Systems
Research with a special focus on responsible recommender systems.
RecSys: Group on Recommender Systems
Our multidisciplinary group works on recommender systems at Tampere University. We focus on various problems, such as:
Recommendations for groups // Sequential recommendations // Fairness, diversity and transparency in recommendations and rankings // Explanations and why-not questions in recommender systems
As well as complementary topics related to the problem of entity resolution, like:
Progressive entity resolution // Iterative blocking and matching // Fairness and diversity in entity resolution // Explanations in entity resolution
We participate in multi-disciplinary research, building on computer science, data science, internet research, social sciences. We aim to design for more ethically sustainable information and communications technology.
RecSys is part of the Faculty of Information Technology and Communication Sciences (ITC) of the Tampere University.
[February 19, 2021] Eirini Ntoutsi from the Leibniz University Hannover and the L3S Research Center in Germany will give the first talk of the year on “Fairness-aware Learning: From Single Models to Sequential Ensemble Learning over Data Streams”. [Video]
[November 20, 2020] Glad to say that Dimitris Sacharidis from the TU Wien will talk about “Efficient Techniques for Ensuring Fairness of Exposure in Recommendations”. [Video]
[October 15, 2020] Maria Stratigi will present us some recent work on why-not questions and explanations for collaborative filtering. [Details]
On September 21, 2020, we restart the Rec Sys Meetings at the Tampere University! Denis Kotkov will talk about conversational recommender systems. Join us on Zoom.
Fairness in Rankings and Recommenders, Tutorial @ EDBT 2020, Joint work with Evaggelia Pitoura and Georgia Koutrika
M. Stratigi, J. Nummenmaa, E. Pitoura and K. Stefanidis, Fair Sequential Group Recommendations, ACM SAC 2020
M. Stratigi, K. Tzompanaki and K. Stefanidis, Why-Not Questions & Explanations for Collaborative Filtering, WISE 2020