Read the original document by opening this link in a new tab.
Table of Contents
1. Abstract
2. Introduction
3. Related Work
4. Method
5. Results
6. Conclusion
7. References
Summary
The document discusses the challenges in traditional sequential recommendation methods and proposes a dual-tower retrieval architecture for sequence recommendation. It introduces a self-supervised multi-modal pretraining method to align various feature combinations of items. The proposed approach aims to address the issue of inconsistent feature levels between model outputs and item embeddings in sequential recommendation. Extensive experiments on five public datasets demonstrate the effectiveness of the method.