Cross-View Topology-Aware Graph Representation Learning
This work addresses graph classification for applications in chemistry, social networks, and bioinformatics, representing an incremental advancement by combining existing methods.
The paper tackled the problem of graph classification by addressing GNNs' oversight of global topological features, proposing GraphTCL, a dual-view contrastive learning framework that integrates structural and topological embeddings, resulting in improved classification performance on benchmark datasets.
Graph classification has gained significant attention due to its applications in chemistry, social networks, and bioinformatics. While Graph Neural Networks (GNNs) effectively capture local structural patterns, they often overlook global topological features that are critical for robust representation learning. In this work, we propose GraphTCL, a dual-view contrastive learning framework that integrates structural embeddings from GNNs with topological embeddings derived from persistent homology. By aligning these complementary views through a cross-view contrastive loss, our method enhances representation quality and improves classification performance. Extensive experiments on benchmark datasets, including TU and OGB molecular graphs, demonstrate that GraphTCL consistently outperforms state-of-the-art baselines. This study highlights the importance of topology-aware contrastive learning for advancing graph representation methods.