FIT CTU Brought the World’s Leading Experts in Recommender Systems to Prague

“Recommender systems are technologies that, based on data analysis, help users discover content they might be interested in – such as movies on streaming platforms, products in e-shops, or articles on news websites,” said Assoc. Prof. Ing. Pavel Kordík, Ph.D., from FIT CTU, co-chair of the conference.

The fact that RecSys 2025 took place in Prague is largely due to the efforts of researchers from FIT CTU and their colleagues from the technology startup Recombee. Together, the two institutions run RecombeeLab, a research lab at FIT CTU that bridges academic research and real-world applications.

“Bringing the world’s largest conference on recommender systems to Prague was a tremendous challenge — and an even greater honor,” said Assoc. Prof. Kordík. “I’m proud that we were able to showcase not only the beauty of our city but also the strength of Czech research and innovation in artificial intelligence,” he added.

Research teams and PhD students from FIT CTU and RecombeeLab contributed several scientific papers to the conference, covering topics such as scalable retrieval in recommender systems, probabilistic modeling and uncertainty estimation, next-basket prediction, and fair group recommendation. They also participated in the INRA 2025 workshop with a notable paper focused on segment-aware analytics for media organizations, inspired by real-world experience from The Telegraph.

“RecSys 2025 in Prague marked a historic milestone – the number of contributions in the industrial track more than doubled compared to previous years, making it the largest in the conference’s history,” said Dr. rer. nat. Rodrigo Augusto da Silva Alves. “The strong connection between academic and applied research in recommender systems proves that at RecombeeLab we’re on the right path, providing our students with unique opportunities for growth,” he added.

As part of the conference’s industrial symposia, experts from leading global companies, including Google DeepMind, gave talks on topics such as scaling inference in large language models and the future impact of quantum computing on recommender systems.

The conference was co-chaired by Assoc. Prof. Ing. Pavel Kordík, Ph.D. (FIT CTU), Prof. Mária Bieliková (Kempelen Institute of Intelligent Technologies), and Prof. Markus Schedl (Johannes Kepler University Linz). The organizing team also included Dr. rer. nat. Rodrigo Augusto da Silva Alves (Assistant Professor, Department of Applied Mathematics, FIT CTU), who led the industrial program, as well as Local Chairs Ing. Vojtěch Vančura, Ph.D. (recent FIT CTU PhD graduate) and Ing. Petr Kasalický (FIT CTU PhD candidate).

Selected presentations, talks, and tutorials from the conference are now publicly available online.

Scientific Contributions

The FIT CTU team presented several original research papers covering key topics in modern recommender systems research:

• The Future is Sparse: Embedding Compression for Scalable Retrieval in Recommender Systems

• Probabilistic Modeling, Learnability and Uncertainty Estimation for Interaction Prediction in Movie Rating Datasets

• Recurrent Autoregressive Linear Model for Next-Basket Recommendation

• SAGEA: Sparse Autoencoder-based Group Embeddings Aggregation for Fairness-Preserving Group Recommendations

• Full-Page Recommender: A Modular Framework for Multi-Carousel Recommendations

FIT CTU and RecombeeLab also contributed to the INRA 2025 workshop with the paper
Segment-Aware Analytics for Real-Time Editorial Support in Media Groups: Lessons from The Telegraph,
which drew on practical insights from the media industry.

More news from FIT