SEGA: Structural Entropy Guided Anchor View for Graph Contrastive Learning
This addresses the issue of semantic alteration in graph contrastive learning for researchers, though it is incremental as it builds on existing graph information bottleneck theory.
The paper tackled the problem of information loss in graph contrastive learning by introducing an anchor view that maintains essential information, achieving significant performance boosts on graph classification benchmarks.
In contrastive learning, the choice of ``view'' controls the information that the representation captures and influences the performance of the model. However, leading graph contrastive learning methods generally produce views via random corruption or learning, which could lead to the loss of essential information and alteration of semantic information. An anchor view that maintains the essential information of input graphs for contrastive learning has been hardly investigated. In this paper, based on the theory of graph information bottleneck, we deduce the definition of this anchor view; put differently, \textit{the anchor view with essential information of input graph is supposed to have the minimal structural uncertainty}. Furthermore, guided by structural entropy, we implement the anchor view, termed \textbf{SEGA}, for graph contrastive learning. We extensively validate the proposed anchor view on various benchmarks regarding graph classification under unsupervised, semi-supervised, and transfer learning and achieve significant performance boosts compared to the state-of-the-art methods.