Research map / Graph learning / GNN architectures
Heterogeneous Graph Structure Learning for Graph Neural Networks
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Learn relation-specific graphs from features, feature propagation, and semantics, then fuse them with the original graph | HGSL node classification · DBLP | Macro-F1: 90.53 -> 91.92 | Helped |
| Learn relation-specific graphs from features, feature propagation, and semantics, then fuse them with the original graph | HGSL node classification · DBLP | Micro-F1: 91.47 -> 92.77 | Helped |
| Learn relation-specific graphs from features, feature propagation, and semantics, then fuse them with the original graph | HGSL node classification · ACM | Macro-F1: 92.96 -> 93.48 | Helped |
| Learn relation-specific graphs from features, feature propagation, and semantics, then fuse them with the original graph | HGSL node classification · ACM | Micro-F1: 92.86 -> 93.37 | Helped |
| Learn relation-specific graphs from features, feature propagation, and semantics, then fuse them with the original graph | HGSL node classification · Yelp | Macro-F1: 93.06 -> 93.55 | Helped |
| Learn relation-specific graphs from features, feature propagation, and semantics, then fuse them with the original graph | HGSL node classification · Yelp | Micro-F1: 92.19 -> 92.76 | Helped |
| Add feature similarity graphs (ablation) | HGSL · DBLP, ACM, and Yelp | Macro-F1 | Helped |
| Add feature propagation graphs (ablation) | HGSL · DBLP, ACM, and Yelp | Macro-F1 | Helped |
| Add semantic graphs (ablation) | HGSL · DBLP, ACM, and Yelp | Macro-F1 | Helped |
| Learn graph-fusion weights with channel attention instead of averaging candidate graphs (ablation) | HGSL · DBLP, ACM, and Yelp | Macro-F1 | Helped |
| Set feature and semantic similarity thresholds to 1, leaving only the original graph structure (ablation) | HGSL · DBLP, ACM, and Yelp | Macro-F1 | Hurt |
Closest papers by meaning
- Heterogeneous Graph Neural Network on Semantic Tree (AAAI 2025 · similarity 0.91)
- HeterSEED: Semantics–Structure Decoupling for Heterogeneous Graph Learning under Heterophily (NeurIPS 2026 · similarity 0.90)
- Bootstrapping Heterogeneous Graph Representation Learning via Large Language Models: A Generalized Approach (AAAI 2025 · similarity 0.90)
- HGOT: Self-supervised Heterogeneous Graph Neural Network with Optimal Transport (ICML 2025 · similarity 0.90)
- NoiseHGNN: Synthesized Similarity Graph-Based Neural Network for Noised Heterogeneous Graph Representation Learning (AAAI 2025 · similarity 0.89)
- An Efficient Subgraph-Inferring Framework for Large-Scale Heterogeneous Graphs (AAAI 2024 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.