Program

Workshop date: September 28, 2026

Time Activity
13:30 - 15:00 Session 1 (90 minutes) — Paper presentations (7 papers)
  1. A RecSys Paper for $20: Why We Must Build, Evaluate, and Govern Autonomous Recommender Systems Research Labs (AutoRecLabs)
    Joeran Beel, Bela Gipp, Tobias Vente, Moritz Baumgart, Philipp Meister and Sinan Pourazari [Experimental Methodologies Track]
  2. Tie Handling Is Part of the Evaluation Protocol: An Order-Invariance Audit for Tie-Heavy Recommender Scores
    Chengkun Guo, Han Chen, Yilin Zhu and Yingrui Li [Experimental Methodologies Track]
  3. Your nDCG Score May Be Correct. But the Conclusion? A Call for Stability-Aware Offline Recommender Evaluation
    Joeran Beel [Experimental Methodologies Track]
  4. The Forgotten Users of Recommender Systems
    Sam Pieters, Basil Rommens, Noah Daniels and Bart Goethals [Experimental Methodologies Track]
  5. Does Diversity Actually Improve Long-term Retention? An Empirical Investigation with Failure Analysis of RL-based Diversity Control
    Hyun Kang and Ji-Hyeong Han [Experimental Methodologies Track]
  6. Calibrating Reproduced Claims in Recommender Systems
    Alan Said [Experimental Methodologies Track]
  7. Serendipity-Aware Recommendation via Candidate Generation and Density-Based Rejection Sampling
    Vivian Wagumba, Joseph Wathome, Mulang Onando and Mwangi Karanja [Research Papers Track]
15:00 - 15:30 Coffee Break — Poster Session (2 papers)
  1. SPADE: Escaping the Popularity-Similarity Frontier to Measure Serendipitous Recommendations
    Tobias Vente, Maarten Peirsman, Noah Daniëls, Hannu Toivonen and Bart Goethals [Experimental Methodologies Track]
  2. Personalizing Diversity in News Recommendation: A User Study
    Aishwarya Satwani and Robin Burke [Research Papers Track]
15:30 - 17:00 Session 2 (90 minutes) — Panel / Interactive Discussion
Panel discussion on concrete ways to improve research evaluation and publication practices in RecSys, with the goal of identifying actionable proposals for the community. Authors of accepted papers are invited to join the discussion.