Bob Junyi Zou

New framework calibrates recommender uncertainty for user retention and diversity

A new research paper introduces a framework for recommender systems that calibrates model uncertainty to improve user experience for both low-activity and high-activity users. For low-active users, the system employs a …