A critical study on data leakage in recommender system offline evaluation

• Main Authors: Ji, Yitong, Sun, Aixin, Zhang, Jie, Li, Chenliang
• Other Authors: School of Computer Science and Engineering
• Format: Article
• Language: English
• Published: 2023
• Subjects: Engineering::Computer science and engineering; Data Mining; Collaborative Filtering
• Online Access: https://hdl.handle.net/10356/170569
• Institution: Nanyang Technological University

Recommender models are hard to evaluate, particularly under offline setting. In this paper, we provide a comprehensive and critical analysis of the data leakage issue in recommender system offline evaluation. Data leakage is caused by not observing global timeline in evaluating recommenders, e.g....