Heterogeneous Graph Condensation via Role-Aware Clustering
This work addresses the computational bottleneck of training HGNNs on large-scale heterogeneous graphs, offering a practical condensation method for practitioners.
HGC-RC proposes a role-aware clustering framework for heterogeneous graph condensation, achieving up to 10x speedup in HGNN training while maintaining or improving accuracy on node classification tasks across multiple datasets.
Heterogeneous Graph Neural Networks (HGNNs) have exhibited remarkable efficacy in modeling complex systems with multiple types of nodes and relations, yet their training on large-scale heterogeneous graphs remains computationally prohibitive. Although graph condensation methods can effectively improve learning efficiency on large-scale graphs, existing condensation processes are mainly designed for homogeneous graphs and typically rely on computationally expensive gradient matching or bilevel optimization paradigms, rendering them impractical for heterogeneous settings. To address these limitations, we propose HGC-RC, a simple yet effective role-aware heterogeneous graph condensation framework. Specifically, HGC-RC first extracts semantically enhanced node embeddings via lightweight propagation. It then introduces a role-aware hybrid clustering strategy consisting of class-partitioned clustering for labeled target nodes to preserve class distributions and unsupervised type-wise clustering for non-target nodes to retain critical cross-type connectivity. Finally, a compact heterogeneous graph is efficiently reconstructed based on the resulting cluster assignments. Extensive experiments demonstrate that HGC-RC outperforms state-of-the-art baselines, offering a practical pathway to accelerate HGNN training on large-scale heterogeneous graphs without sacrificing task performance