Rethinking Multi-Label Node Classification: Do Tuned Classic GNNs Suffice?

TL;DR

This paper demonstrates that carefully tuned classic GNNs can outperform specialized multi-label node classification methods, emphasizing the importance of strong baselines in evaluating new models.

Contribution

It shows that optimized traditional GNNs serve as strong baselines for MLNC, challenging the necessity of complex specialized designs.

Findings

Tuned GCN, SSGConv, and GCNII outperform specialized methods on most datasets.

Careful optimization of classic GNNs achieves state-of-the-art results.

Highlights the importance of strong baselines in MLNC research.

Abstract

Multi-label node classification (MLNC) has recently been addressed by increasingly complex label-aware designs that explicitly model node-label interactions and inter-label dependencies.However, it remains unclear whether the advantages of these methods truly stem from their specialized designs, or simply from insufficiently optimized baselines. In this paper, we revisit MLNC from a strong-baseline perspective and investigate whether carefully tuned classic full-graph GNNs can already serve as strong solutions to this task. We systematically study several representative backbones, including GCN, SSGConv, and GCNII, and optimize them using standard yet effective techniques such as normalization, dropout, and residual connections. Experiments on five representative benchmark datasets show that our tuned baselines outperform representative specialized methods on four datasets and achieve state-of-the-art performance in multiple settings. These results indicate that careful tuning of classic backbones is a highly influential but often overlooked factor in MLNC, and highlight the need for more rigorous strong-baseline evaluation in future research on multi-label graph learning.