GraphGini

code for paper: "GraphGini: Fostering Individual and Group Fairness in Graph Neural Networks"

1. Installation Setup

Please run the following commands to install necessary packages. For more details on Pytorch Geometric please refer to install the PyTorch Geometric packages following the instructions from here.

conda create --name graphgini python==3.7.11
conda activate graphgini
conda install pytorch==1.10.0 cudatoolkit=11.1 -c pytorch -c conda-forge

pip install torch-scatter -f https://data.pyg.org/whl/torch-1.10.0+cu111.html
pip install torch-sparse -f https://data.pyg.org/whl/torch-1.10.0+cu111.html
pip install torch-geometric==2.0.1

pip install aif360==0.3.0

2. Datasets

We ran our experiments on two high-stakes datasets: credit and income and also a larger social network dataset pokec. All the data are present in the './dataset' folder. Due to space constraints, pokec is zipped so please unzip it before use.

3. Use bellow commands to run algorithms

cd code
python GraphGini.py  --model [gcn/gin/jk] --dataset [DATASET_NAME] --epochs [number_of_epochs]

For running GraphGini without GradNorm i.e (\beta_1 = 1, \beta_2 = alpha, \beta_3 = beta)

cd code
python GraphGini_WGN.py  - --model gcn --dataset credit --alpha 6e-6 --beta 2

4. Baselines

We used code provided by "https://github.com/michaelweihaosong/GUIDE.git" to run all baselines