Vector Quantization for Recommender Systems: A Review and Outlook

TL;DR

This paper reviews vector quantization techniques in recommender systems, discussing their applications, challenges, and future trends, especially in the context of large models and multimodal data, to guide future research.

Contribution

It provides a comprehensive taxonomy and analysis of vector quantization methods for recommender systems, highlighting challenges and future directions in the field.

Findings

Vector quantization enhances feature compression in recommender systems.

Challenges include training difficulties and integration with large language models.

Emerging trends involve multimodal data and advanced training techniques.

Abstract

Vector quantization, renowned for its unparalleled feature compression capabilities, has been a prominent topic in signal processing and machine learning research for several decades and remains widely utilized today. With the emergence of large models and generative AI, vector quantization has gained popularity in recommender systems, establishing itself as a preferred solution. This paper starts with a comprehensive review of vector quantization techniques. It then explores systematic taxonomies of vector quantization methods for recommender systems (VQ4Rec), examining their applications from multiple perspectives. Further, it provides a thorough introduction to research efforts in diverse recommendation scenarios, including efficiency-oriented approaches and quality-oriented approaches. Finally, the survey analyzes the remaining challenges and anticipates future trends in VQ4Rec, including the challenges associated with the training of vector quantization, the opportunities presented by large language models, and emerging trends in multimodal recommender systems. We hope this survey can pave the way for future researchers in the recommendation community and accelerate their exploration in this promising field.