GAEA

[PAKDD2023] The source codes for Improving Knowledge Graph Entity Alignment with Graph Augmentation.

Getting Started

Datasets

We use entity alignment benchmark datasets OpenEA which can be downloaded from OpenEA. You need to put the prepared data into ../data/ folder.

Dependencies

• Python 3
• PyTorch
• networkx==2.5.1
• Scipy
• Numpy
• Pandas
• Scikit-learn

You can automatically download corresponding dependencies by following scripts:

conda create -n GAEA python=3.6
conda activate GAEA
conda install -n GAEA pytorch=1.10.2 torchvision torchaudio cudatoolkit=11.3.1 -c pytorch # change according to your need here
pip install -r .\requirements.txt

Running

To run GAEA, please use the following scripts (ps: --task is an argument):

python train.py --task en_fr_15k
python train.py --task en_de_15k
python train.py --task d_w_15k
python train.py --task d_y_15k

To run 5-fold cross-validation, please use the following script:

python run_fold.py --task en_fr_15k

We also provide jupyter notebook version in GAEA.ipynb.

If you have any difficulty or question in running code and reproducing experimental results, please email to xiefeng@nudt.edu.cn.

Acknowledgement

We refer to the codes of these repos: GCN-Align, OpenEA, MuGNN, IMEA. Thanks for their great contributions!