Ego-GNNs: Exploiting Ego Structures in Graph Neural Networks
| Autor: | William L. Hamilton, Dylan Sandfelder, Priyesh Vijayan |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
FOS: Computer and information sciences
Computer Science - Machine Learning Transitive relation Theoretical computer science Computer science Inductive bias business.industry Node (networking) Deep learning Perspective (graphical) Induced subgraph Machine Learning (cs.LG) Convolution Kernel (image processing) Artificial intelligence business |
| Zdroj: | ICASSP |
| Popis: | Graph neural networks (GNNs) have achieved remarkable success as a framework for deep learning on graph-structured data. However, GNNs are fundamentally limited by their tree-structured inductive bias: the WL-subtree kernel formulation bounds the representational capacity of GNNs, and polynomial-time GNNs are provably incapable of recognizing triangles in a graph. In this work, we propose to augment the GNN message-passing operations with information defined on ego graphs (i.e., the induced subgraph surrounding each node). We term these approaches Ego-GNNs and show that Ego-GNNs are provably more powerful than standard message-passing GNNs. In particular, we show that Ego-GNNs are capable of recognizing closed triangles, which is essential given the prominence of transitivity in real-world graphs. We also motivate our approach from the perspective of graph signal processing as a form of multiplex graph convolution. Experimental results on node classification using synthetic and real data highlight the achievable performance gains using this approach. Submitted to a special session of IEEE-ICASSP 2021 |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|