Hattrick at A-rated UMAP 2026
Three full papers, co-authored by EXDIGIT professor Christine Bauer, have been accepted at the upcoming A-rated UMAP 2026 conference (User Modeling, Adaptation, and Personalization). These works span PhD student-led work with Utrecht University and an industry collaboration with DPG Media.
Key themes
• PhD student-led research on modeling preference heterogeneity:
Led by Anouk van Kasteren, co-supervised with Judith Masthoff, and in collaboration with PhD candidate Marloes Vredenborg (all Utrecht University), this paper models preference heterogeneity for context-aware decision support during public transport disruptions, exploring how situational factors shape preferences and choices under real-world conditions.
• PhD student-led study on tone in recommendation explanations: Led by Marloes Vredenborg, co-supervised with Judith Masthoff and Marit Bentvelzen (all Utrecht University), this paper shows how situational context—especially urgency—shapes user preferences for explanation tone. It introduces a generalizable methodology for studying and designing context- and tone-aware explanations and, in the public-transport domain, offers qualitative insights into users’ needs and a validated set of explanation tones.
• Industry collaboration (DPG Media) on news recommender diversification: Co-authored with Robin Verachtert and Kim Falk (DPG Media), this paper reports a large-scale comparison of three diversification methods deployed on NU.nl (a Dutch online newspaper)—interleaving and two intra-list diversification (ILD) variants (TF-IDF, BERT). In the deployment on NU.nl, ILD with BERT reduced intra-list similarity, increased click-through rates, and improved perceived relevance—aligning diversification with editorial goals.
Learn more
UMAP 2026: https://www.um.org/umap2026/
