LTGNN-PyTorch

This is the Pytorch implementation for our TheWebConf'24 paper "Linear-Time Graph Neural Networks for Scalable Recommendations". Please find our paper in ACM Digital Library (https://dl.acm.org/doi/10.1145/3589334.3645486) or arXiv (https://arxiv.org/abs/2402.13973).

The Amazon-Large dataset is available on Google Drive. If you prefer to build it yourself, please refer to Section 4.1 of our paper for instructions on constructing it directly from the Amazon Review Data.

This codebase was adapted from LightGCN-pytorch.

Enviroment Requirement

pip install -r requirements.txt

Command

cd code && python main.py --decay=2e-4 --lr=0.0015 --layer=1 --seed=2020 --dataset="yelp2018" --topks="[20]" --recdim=64 --model="ltgnn" --appnp_alpha=0.45 --num_neighbors=15 --device=0

cd code && python main.py --decay=2e-4 --lr=0.0015 --layer=1 --seed=2020 --dataset="alibaba-ifashion" --topks="[20]" --recdim=64 --model="ltgnn" --appnp_alpha=0.45 --num_neighbors=15 --device=0

BibTeX

If you find LTGNN useful in your research, please cite the following in your manuscript:

@inproceedings{zhang2024linear,
  title={Linear-Time Graph Neural Networks for Scalable Recommendations},
  author={Zhang, Jiahao and Xue, Rui and Fan, Wenqi and Xu, Xin and Li, Qing and Pei, Jian and Liu, Xiaorui},
  booktitle={Proceedings of the ACM on Web Conference 2024},
  pages = {3533-3544},
  year={2024}
}