2024/08/26 by Nikhil Khani, Khani, Nikhil, Shuo Yang +21
Computer Science · Decision Sciences · Economics, Econometrics and Finance · #Artificial Intelligence (cs.AI) #Auction Theory and Applications #FOS: Computer and information sciences #Game Theory and Voting Systems #Information Retrieval (cs.IR) #Machine Learning (cs.LG) #Spam and Phishing Detection
paper · pdf · doi:10.48550/arxiv.2408.14678
openalex publication_date 2024/08/26 · openalex created_date 2024/09/21 · openalex updated_date 2026/07/28
Knowledge Distillation (KD) is a powerful approach for compressing a large model into a smaller, more efficient model, particularly beneficial for latency-sensitive applications like recommender systems. However, current KD research predominantly focuses on Computer Vision (CV) and NLP tasks, overlooking unique data characteristics and challenges inherent to recommender systems. This paper addresses these overlooked challenges, specifically: (1) mitigating data distribution shifts between teacher and student models, (2) efficiently identifying optimal teacher configurations within time and budgetary constraints, and (3) enabling computationally efficient and rapid sharing of teacher labels to support multiple students. We present a robust KD system developed and rigorously evaluated on multiple large-scale personalized video recommendation systems within Google. Our live experiment results demonstrate significant improvements in student model performance while ensuring consistent and reliable generation of high quality teacher labels from a continuous data stream of data.