DGRec

A PyTorch and DGL implementation for the WSDM 2023 paper below:
DGRec: Graph Neural Network for Recommendation with Diversified Embedding Generation

Environment DGL version 1.0.1 Pytorch version 1.12.1

Running

python main.py
Then you can get similar result on TaoBao dataset as illustrated in the paper.

You can check different hyper-parameters in utils/parser.py

Dataset

Format of train.txt val.txt test.txt: UserID,ItemID.

Format of item_category.txt: ItemID,CategoryID

If you want to change to your own dataset. Format your dataset in a folder as in datasets/Beauty, and specify your folder name in --dataset argument.

Citation

If you use our code, please cite the paper below:

@inproceedings{yang2023dgrec,
  title={DGRec: Graph Neural Network for Recommendation with Diversified Embedding Generation},
  author={Yang, Liangwei and Wang, Shengjie and Tao, Yunzhe and Sun, Jiankai and Liu, Xiaolong and Yu, Philip S and Wang, Taiqing},
  booktitle={Proceedings of the Sixteenth ACM International Conference on Web Search and Data Mining},
  pages={661--669},
  year={2023}
}