The authors address the challenge of providing personalized insurance recommendations in a data-poor domain by developing innovative recommender models that outperform existing baselines.
