Augmentation-Free Self-Supervised Learning on Graphs

TL;DR

This paper introduces AFGRL, an augmentation-free self-supervised learning framework for graphs that leverages structural and semantic similarities, outperforming augmentation-based methods across multiple node-level tasks.

Contribution

The paper proposes a novel augmentation-free approach for graph self-supervised learning, reducing dependency on augmentation schemes and improving performance.

Findings

AFGRL outperforms existing augmentation-based methods on node classification.

The method is effective across clustering and similarity search tasks.

Extensive experiments validate the superiority of AFGRL on real-world datasets.

Abstract

Inspired by the recent success of self-supervised methods applied on images, self-supervised learning on graph structured data has seen rapid growth especially centered on augmentation-based contrastive methods. However, we argue that without carefully designed augmentation techniques, augmentations on graphs may behave arbitrarily in that the underlying semantics of graphs can drastically change. As a consequence, the performance of existing augmentation-based methods is highly dependent on the choice of augmentation scheme, i.e., hyperparameters associated with augmentations. In this paper, we propose a novel augmentation-free self-supervised learning framework for graphs, named AFGRL. Specifically, we generate an alternative view of a graph by discovering nodes that share the local structural information and the global semantics with the graph. Extensive experiments towards various node-level tasks, i.e., node classification, clustering, and similarity search on various real-world datasets demonstrate the superiority of AFGRL. The source code for AFGRL is available at https://github.com/Namkyeong/AFGRL.