Robustness of Graph Self-Supervised Learning to Real-World Noise: A Case Study on Text-Driven Biomedical Graphs

TL;DR

This paper evaluates the robustness of Graph Self-Supervised Learning methods on noisy, text-derived biomedical graphs, proposing a framework that improves performance and offers practical guidance for real-world applications.

Contribution

It introduces NATD-GSSL, a comprehensive framework for GSSL on noisy biomedical graphs, and provides the first systematic robustness analysis in this context.

Findings

Relation reconstruction is highly sensitive to noise but benefits from schemas.

Feature reconstruction remains robust and comparable to clean graphs.

Bidirectional GNNs outperform unidirectional ones on noisy graphs.

Abstract

Graph Self-Supervised Learning (GSSL) offers a powerful paradigm for learning graph representations without labeled data. However, existing work assumes clean, manually curated graphs. Recent advances in NLP enable the large-scale automatic extraction of knowledge graphs from text, opening new opportunities for GSSL while introducing substantial real-world noise. This type of noise remains largely unexplored, as prior robustness studies typically rely on synthetic perturbations. To address this gap, we present the first comprehensive evaluation of GSSL methods on text-driven graphs for unsupervised term typing. We introduce Noise-Aware Text-Driven Graph GSSL (NATD-GSSL), a unified framework that combines automatic graph construction, graph refinement, and GSSL. Our evaluation follows a dual-graph protocol that contrasts a noisy graph derived from MedMentions with a clean Unified Medical Language System (UMLS) reference graph, aligned through a shared gold standard. Our results reveal variability in robustness across both pretext tasks and Graph Neural Network (GNN) architectures. Relation reconstruction is highly sensitive to noise and benefits from well-defined schemas, whereas feature reconstruction is considerably more robust, achieving performance comparable to clean-graph settings. Contrastive objectives are generally less affected by noise but depend strongly on alignment with downstream tasks. GNN architecture also plays a critical role: bidirectional relational message-passing designs are better suited to noisy, text-driven graphs, while unidirectional relational ones perform best on clean graphs. Overall, NATD-GSSL provides practical guidance for applying GSSL to real-world, noisy graphs and achieves up to a 7\% improvement over pretrained language model baselines. All code and benchmarks are publicly available at https://github.com/OthmaneKabal/MC2GAE.