Tag
This research fine-tunes LLMs on human survey data to serve as judgmental models for group recommender systems, dynamically selecting aggregation strategies to maximize satisfaction and consensus. A user study validates that the approach aligns with human fairness and satisfaction perceptions.