Towards a Unified Framework for Fair and Stable Graph Representation Learning
This work addresses fairness and stability issues in graph representation learning, which is crucial for high-stakes applications like criminal justice and financial lending, though it appears incremental as it builds on existing GNN methods.
The authors tackled the problem of ensuring fairness and stability in graph neural network representations by proposing NIFTY, a framework that unifies counterfactual fairness and stability, showing theoretical and experimental improvements on new datasets in criminal justice and financial lending domains.
As the representations output by Graph Neural Networks (GNNs) are increasingly employed in real-world applications, it becomes important to ensure that these representations are fair and stable. In this work, we establish a key connection between counterfactual fairness and stability and leverage it to propose a novel framework, NIFTY (uNIfying Fairness and stabiliTY), which can be used with any GNN to learn fair and stable representations. We introduce a novel objective function that simultaneously accounts for fairness and stability and develop a layer-wise weight normalization using the Lipschitz constant to enhance neural message passing in GNNs. In doing so, we enforce fairness and stability both in the objective function as well as in the GNN architecture. Further, we show theoretically that our layer-wise weight normalization promotes counterfactual fairness and stability in the resulting representations. We introduce three new graph datasets comprising of high-stakes decisions in criminal justice and financial lending domains. Extensive experimentation with the above datasets demonstrates the efficacy of our framework.