The code of our paper "RaSeRec: Retrieval-Augmented Sequential Recommendation" [pdf].
We have provided checkpoints trained on the Beauty datasets under the log directory.
We propose a new SeRec learning paradigm, RaSeRec, which explores RAG in sequential recommendation (SeRec) to solve issues existing in previous paradigms, i.e., preference drift and implicit memory.
We have provided the Beatuty dataset. More datasets can be downloaded from RecSysDatasets or their Google Drive. And put the files in ./dataset/ like the following.
$ tree
├── Amazon_Beauty
├── Amazon_Beauty.inter
└── Amazon_Beauty.item
Run raserec.sh.
If you find this repo useful, please cite
@misc{zhao2024raserec,
title={RaSeRec: Retrieval-Augmented Sequential Recommendation},
author={Xinping Zhao and Baotian Hu and Yan Zhong and Shouzheng Huang and Zihao Zheng and Meng Wang and Haofen Wang and Min Zhang},
year={2024},
eprint={2412.18378},
archivePrefix={arXiv},
primaryClass={cs.IR}
}