Abstract
Graph Neural Networks (GNNs) have been widely adopted for graph-structured data analysis. However, the performance of standard GNN models often suffers when applied to heterophilous graphs, where connected nodes tend to have dissimilar labels. This paper introduces the Contextualized Graph Neural Network (CGNN), a novel GNN architecture that enhances node representation learning by explicitly modeling the context of each node. CGNN uses a context encoding module to capture contextual information, followed by an adaptive message-passing mechanism, and finally a context-aware aggregation module. We provide a detailed algorithm and a Python implementation using PyTorch. Experimental results on benchmark heterophilous graph datasets demonstrate that CGNN outperforms standard GNNs and state-of-the-art heterophilyfocused methods.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
"Contextualized Graph Neural Network for Node Classification", Technical Disclosure Commons, (February 03, 2025)
https://www.tdcommons.org/dpubs_series/7801