Graph Neural Distance Metric Learning with Graph-Bert

2020/02/09 by Jiawei Zhang, Zhang, Jiawei
Biochemistry, Genetics and Molecular Biology · Computer Science · #Advanced Graph Neural Networks #Epigenetics and DNA Methylation #Graph Theory and Algorithms

paper · pdf · doi:10.48550/arxiv.2002.03427

openalex publication_date 2020/02/09 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Graph distance metric learning serves as the foundation for many graph learning problems, e.g., graph clustering, graph classification and graph matching. Existing research works on graph distance metric (or graph kernels) learning fail to maintain the basic properties of such metrics, e.g., non-negative, identity of indiscernibles, symmetry and triangle inequality, respectively. In this paper, we will introduce a new graph neural network based distance metric learning approaches, namely GB-DISTANCE (GRAPH-BERT based Neural Distance). Solely based on the attention mechanism, GB-DISTANCE can learn graph instance representations effectively based on a pre-trained GRAPH-BERT model. Different from the existing supervised/unsupervised metrics, GB-DISTANCE can be learned effectively in a semi-supervised manner. In addition, GB-DISTANCE can also maintain the distance metric basic properties mentioned above. Extensive experiments have been done on several benchmark graph datasets, and the results demonstrate that GB-DISTANCE can out-perform the existing baseline methods, especially the recent graph neural network model based graph metrics, with a significant gap in computing the graph distance.

Citations

• Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling
• Long Short-Term Memory
• Semi-Supervised Classification with Graph Convolutional Networks
• Junction Tree Variational Autoencoder for Molecular Graph Generation
• IsoNN: Isomorphic Neural Network for Graph Representation Learning and\n Classification
• GResNet: Graph Residual Network for Reviving Deep GNNs from Suspended Animation
• Graph-Bert: Only Attention is Needed for Learning Graph Representations

Related

• Distance Metric Learning using Graph Convolutional Networks: Application\n to Functional Brain Networks
• Learning Graph Edit Distance by Graph Neural Networks
• A Simple Way to Learn Metrics Between Attributed Graphs
• Distance Encoding: Design Provably More Powerful Neural Networks for Graph Representation Learning
• GREED: A Neural Framework for Learning Graph Distance Functions