Node Classification With Integrated Reject Option
Uday Bhaskar, Jayadratha Gayen, Charu Sharma, Naresh Manwani
TL;DR
NCwR introduces two abstention-aware graph neural network architectures for node classification: NCwR-Cov (coverage-based) and NCwR-Cost (cost-based). By treating abstention as a controlled decision via a selective predictor or an extra rejection class, they outperform Softmax-Response and conformal-prediction baselines across standard citation datasets and ILDC-based legal judgment prediction tasks. The approach uses a GAT backbone, calibrated selection thresholds or cost parameters, and SHAP-based explanations to illuminate abstentions. Overall, integrating reject options in GNNs enhances reliability in high-stakes domains like law and healthcare, while remaining broadly applicable and explainable.
Abstract
One of the key tasks in graph learning is node classification. While Graph neural networks have been used for various applications, their adaptivity to reject option setting is not previously explored. In this paper, we propose NCwR, a novel approach to node classification in Graph Neural Networks (GNNs) with an integrated reject option, which allows the model to abstain from making predictions when uncertainty is high. We propose both cost-based and coverage-based methods for classification with abstention in node classification setting using GNNs. We perform experiments using our method on three standard citation network datasets Cora, Citeseer and Pubmed and compare with relevant baselines. We also model the Legal judgment prediction problem on ILDC dataset as a node classification problem where nodes represent legal cases and edges represent citations. We further interpret the model by analyzing the cases that the model abstains from predicting by visualizing which part of the input features influenced this decision.