Self-Supervised Graph Learning via Spectral Bootstrapping and Laplacian-Based Augmentations

TL;DR

LaplaceGNN introduces a spectral bootstrapping approach for self-supervised graph learning that avoids negative sampling and handcrafted augmentations, leading to efficient and robust graph representations.

Contribution

It proposes a novel spectral augmentation and adversarial training scheme that enhances self-supervised graph learning without contrastive objectives.

Findings

Outperforms state-of-the-art self-supervised graph methods

Achieves linear scaling and efficiency

Provides rich structural representations

Abstract

We present LaplaceGNN, a novel self-supervised graph learning framework that bypasses the need for negative sampling by leveraging spectral bootstrapping techniques. Our method integrates Laplacian-based signals into the learning process, allowing the model to effectively capture rich structural representations without relying on contrastive objectives or handcrafted augmentations. By focusing on positive alignment, LaplaceGNN achieves linear scaling while offering a simpler, more efficient, self-supervised alternative for graph neural networks, applicable across diverse domains. Our contributions are twofold: we precompute spectral augmentations through max-min centrality-guided optimization, enabling rich structural supervision without relying on handcrafted augmentations, then we integrate an adversarial bootstrapped training scheme that further strengthens feature learning and robustness. Our extensive experiments on different benchmark datasets show that LaplaceGNN achieves superior performance compared to state-of-the-art self-supervised graph methods, offering a promising direction for efficiently learning expressive graph representations.