This repository is the PyTorch impelementation for WSDM 2024 paper:
Linear Recurrent Units for Sequential Recommendation [Paper][Code] (BibTex citation at the bottom)
Zhenrui Yue*, Yueqi Wang*, Zhankui He†, Huimin Zeng, Julian McAuley, Dong Wang. Linear Recurrent Units for Sequential Recommendation.
Numpy, pandas, pytorch etc. For our detailed running environment see requirements.txt
The command below specifies the training of LRURec on MovieLens-1M.
python train.py --dataset_code=ml-1mExcecute the above command (with arguments) to train LRURec, select dataset_code from ml-1m, beauty, video, sports, steam and xlong. XLong must be downloaded separately and put under ./data/xlong for experiments. Once trainin is finished, evaluation is automatically performed with models and results saved in ./experiments.
The table below reports our main performance results, with best results marked in bold and second best results underlined. For training and evaluation details, please refer to our paper.
Please consider citing the following paper if you use our methods in your research:
@inproceedings{yue2024linear,
author={Yue, Zhenrui and Wang, Yueqi and He, Zhankui and Zeng, Huimin and McAuley, Julian and Wang, Dong},
title = {Linear Recurrent Units for Sequential Recommendation},
year = {2024},
booktitle = {Proceedings of the Seventeenth ACM International Conference on Web Search and Data Mining},
series = {WSDM '24}
}