END4Rec is an efficient noise-decoupling approach designed for multi-behavior sequential recommendation. It introduces a Noise-Decoupling Mechanism to better model user interactions, reduce noise interference, and improve recommendation performance. The method is particularly useful for large-scale recommendation datasets such as CIKM, IJCAI, and Taobao.
- CIKM (Link): Contains multi-behavior user interactions (e.g., clicks, purchases) for sequential recommendation research.
- IJCAI (Link): A real-world dataset from Taobao, incorporating various user-item interactions.
- Taobao (Link): A large-scale e-commerce dataset covering user shopping behavior.
The MindSpore implementation of END4Rec is optimized for parallel computation and efficient training. Since there is a large amount of framework-specific configuration code, please focus primarily on the implementation of the model. If you have any questions, please contact WeChat: 18331556501.