$\mathcal{G}^2Pxy$: Generative Open-Set Node Classification on Graphs with Proxy Unknowns

TL;DR

This paper introduces $ ext{G}^2 ext{Pxy}$, a generative open-set node classification method that uses proxy unknowns and mixup to improve detection of unseen classes in graphs without prior unknown class data.

Contribution

The paper proposes a novel inductive open-set classification approach with proxy unknown nodes generated via mixup, enhancing unknown class detection without prior unknown data.

Findings

Outperforms existing methods on benchmark datasets.

Effectively detects unknown classes with high accuracy.

Compatible with various GNN architectures.

Abstract

Node classification is the task of predicting the labels of unlabeled nodes in a graph. State-of-the-art methods based on graph neural networks achieve excellent performance when all labels are available during training. But in real-life, models are often applied on data with new classes, which can lead to massive misclassification and thus significantly degrade performance. Hence, developing open-set classification methods is crucial to determine if a given sample belongs to a known class. Existing methods for open-set node classification generally use transductive learning with part or all of the features of real unseen class nodes to help with open-set classification. In this paper, we propose a novel generative open-set node classification method, i.e. $\mathcal{G}^2Pxy$, which follows a stricter inductive learning setting where no information about unknown classes is available during training and validation. Two kinds of proxy unknown nodes, inter-class unknown proxies and external unknown proxies are generated via mixup to efficiently anticipate the distribution of novel classes. Using the generated proxies, a closed-set classifier can be transformed into an open-set one, by augmenting it with an extra proxy classifier. Under the constraints of both cross entropy loss and complement entropy loss, $\mathcal{G}^2Pxy$ achieves superior effectiveness for unknown class detection and known class classification, which is validated by experiments on benchmark graph datasets. Moreover, $\mathcal{G}^2Pxy$ does not have specific requirement on the GNN architecture and shows good generalizations.