Mitigating popularity bias in recommendation with unbalanced interactions: A gradient perspective

• Main Authors: REN, Weijieying, WANG, Lei, LIU, Kunpeng, GUO, Ruocheng, LIM, Ee-peng, FU, Yanjie
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2022
• Subjects: Training; Calibration; Data mining; Task analysis; Recommender systems; Optimization; Testing; Databases and Information Systems; Data Storage Systems
• Online Access: https://ink.library.smu.edu.sg/sis_research/7510; https://ink.library.smu.edu.sg/context/sis_research/article/8513/viewcontent/2211.01154.pdf
• Institution: Singapore Management University

Recommender systems learn from historical user-item interactions to identify preferred items for target users. These observed interactions are usually unbalanced following a long-tailed distribution. Such long-tailed data lead to popularity bias to recommend popular but not personalized items to use...