• Compatible with PyTorch 1.8 and Python 3.8.
• Dependencies can be installed using requirements.txt.

• Make directories: ./checkpoints, ./log, ./wandb

• We conduct experiments on three datasets: MAG-G, Aminer-G, and Weeplaces-G. The datasets can be downloaded here
• The datasets should be put in ./data directory.

• Params for reproducing the reported GE+MMAN results are saved in ./config. Please run

    #### Method: GE+MMAN
    # MAG-G
    python run_academic_group.py -config_file MAG-G.yml
    
    # Aminer-G
    python run_academic_group.py -config_file Aminer-G.yml
    
    # Weeplaces-G
    python run_weeplaces.py -config_file Weeplaces-G.yml

• Commands for reproducing the other methods (take MAG-G dataset as a example):

  # -----------Plain GNN-based Encoders----------- #
  # GCN
  python run_academic_group.py -dataset MAG-G -only_GE True -graph_based GraphSage -score_method MLP
  
  # -----------GE with Aggregating Methods----------- #
  # GE-AVG
  python run_academic_group.py -dataset MAG-G -only_GE False -graph_based GraphSage -i2g_method avg -score_method MLP
  
  # GE-MMAN
  python run_academic_group.py -dataset MAG-G -only_GE False -graph_based GraphSage -i2g_method MMAN -view_num 4 -score_method mv_score

• -only_GE denotes only using the GNN-based Encoders without aggregators.
• -graph_based denotes GNN encoder type.
• -i2g_method denotes the aggragating method. It can take the following values:
  • avg/degree/att/set2set/MMAN/MAB
• -score_method is some name given for the run (used for storing model parameters)
  • mv_score for MMAN
  • MLP for other aggragating methods
• Rest of the arguments can be listed using python run_academic_group.py -h or python run_weeplaces.py -h

Please cite the following paper if you use this code in your work.

@inproceedings{
    Yin2023beyondindividuals,
    title={Beyond Individuals: Modeling Mutual and Multiple Interactions for Inductive Link Prediction between Groups},
    author={Gongzhu Yin and Xing Wang and Hongli Zhang and Chao Meng and Yuchen Yang and Ku Lu and Yi Luo},
    booktitle={ACM International Conference on Web Search and Data Mining},
    year={2023}
}

For any questions or suggestions please contact Gongzhu Yin ([email protected]).