In this project, we open-source the source code of our GNN-PE approach, including both offline and online processes (i.e., offline.tar.gz and online.tar.gz).
The real-world and default synthetic datasets used in our paper are stored in the datasets directory. As some synthetic datasets are large, we do not upload them. You can easily generate them by following the instruction in our paper.
On Git Hub, we will introduce how to reproduce the results of our experiments over the Yeast dataset.
Note that since we have prepared all the necessary data, you can run the online process directly without running offline process first.
tar -xzvf offline.tar.gz
cd offline
conda create --name <new_environment_name> --file requirements.txt
conda activate <new_environment_name>
python main.py
tar -xzvf online.tar.gz
cd online
mkdir build
cd build
cmake ..
make
./build/src/main