Achieving Group and Individual Fairness in Graph Neural Networks
This paper proposes a novel framework named FairGI that simultaneously achieves group fairness and individual fairness within groups in graph learning, while maintaining comparable prediction performance.