Hierarchical Time-Aware Mixture of Experts for Multi-Modal Sequential Recommendation (WWW2025)

This is the official pytorch implementation of our paper "Hierarchical Time-Aware Mixture of Experts for Multi-Modal Sequential Recommendation", which is accepted by WWW2025.

Running Instruction

To run HM4SR, just use the following code.

python run_hm4sr.py

We provide the processed Games dataset link for your reference. After downloading, just unzip the zip file into the folder of HM4SR.

Reference

@inproceedings{zhang2025hierarchical,
  title={Hierarchical Time-Aware Mixture of Experts for Multi-Modal Sequential Recommendation},
  author={Zhang, Shengzhe and Chen, Liyi and Shen, Dazhong and Wang, Chao and Xiong, Hui},
  booktitle={Proceedings of the ACM on Web Conference 2025},
  pages={3672--3682},
  year={2025}
}

Acknowledgement

Our implementation is based on Recbole. We also referred to the codes of UniSRec. Thanks for the splendid codes for these authors.