CSGNN: Conquering Noisy Node labels via Dynamic Class-wise Selection

TL;DR

CSGNN introduces a dynamic class-wise selection method that uses neighbor-aggregated confidences to identify clean nodes, effectively addressing label noise and class imbalance in graph neural networks, leading to improved robustness and performance.

Contribution

The paper proposes CSGNN, a novel class-wise selection approach that adaptively identifies reliable nodes using neighbor-aggregated confidences, improving GNN robustness against noisy labels and class imbalance.

Findings

CSGNN outperforms state-of-the-art methods in effectiveness.

CSGNN demonstrates enhanced robustness to label noise.

The method effectively handles class imbalance issues.

Abstract

Graph Neural Networks (GNNs) have emerged as a powerful tool for representation learning on graphs, but they often suffer from overfitting and label noise issues, especially when the data is scarce or imbalanced. Different from the paradigm of previous methods that rely on single-node confidence, in this paper, we introduce a novel Class-wise Selection for Graph Neural Networks, dubbed CSGNN, which employs a neighbor-aggregated latent space to adaptively select reliable nodes across different classes. Specifically, 1) to tackle the class imbalance issue, we introduce a dynamic class-wise selection mechanism, leveraging the clustering technique to identify clean nodes based on the neighbor-aggregated confidences. In this way, our approach can avoid the pitfalls of biased sampling which is common with global threshold techniques. 2) To alleviate the problem of noisy labels, built on the concept of the memorization effect, CSGNN prioritizes learning from clean nodes before noisy ones, thereby iteratively enhancing model performance while mitigating label noise. Through extensive experiments, we demonstrate that CSGNN outperforms state-of-the-art methods in terms of both effectiveness and robustness.