We visited the 32nd ACM Conference on User Modeling, Adaptation and Personalization. ACM UMAP is the premier international conference for researchers and practitioners working on systems that adapt to individual users, to groups of users, and that collect, represent, and model user information. The cpnference happened between 1st and 4th July at University of Cagliari.
Students from HICUP Lab presented two doctoral consortium papers:
- Arsen Matej Golubovikj, Improving Recommendations for Non-Mainstream Users by Addressing Subjective Item Views, https://doi.org/10.1145/3631700.3664916
This research addresses the challenge of improving recommender systems to better serve non-mainstream users whose preferences differ from popular trends, aiming to reduce existing biases. By aligning item representations with users’ subjective perceptions, it seeks to enhance recommendation accuracy, especially for users with niche or uncommon tastes. - Uroš Sergaš, Perspective Diversification News Recommender System, https://doi.org/10.1145/3631700.3664918
This research aims to develop a news recommender system designed to reduce political polarization by gradually presenting diverse viewpoints within the Overton window, fostering critical thinking without attempting to persuade users to change their beliefs. By evaluating existing methods and creating a prototype for depolarization, the study provides valuable insights and frameworks for addressing the negative impacts of political polarization on mental health and social cohesion, with significant potential for future research in this area.
