Research map / Graph learning / GNN architectures
HC2-GNN: Hierarchical Graph Representation Learning for Efficient Text Classification
jiejie fan, Xiaojuan Ban, Zhiyan Zhang, Xi Sun
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| add C2GC clustering and coarsening (ablation) | HC2-GNN · 20NG | accuracy (%): 86.5 -> 89.6 | Improved |
| add text-order-aware clustering instead of topology-only conductance (ablation) | HC2-GNN · 20NG-L | accuracy (%): 90.7 -> 94.3 | Improved |
| add text-order-aware clustering instead of topology-only conductance (ablation) | HC2-GNN · 20NG | accuracy (%): 87.9 -> 89.6 | Improved |
| add virtue cluster extension (ablation) | HC2-GNN · 20NG | accuracy (%): 88.4 -> 89.6 | Improved |
| add virtue cluster extension (ablation) | HC2-GNN · 20NG-L | accuracy (%): 92.6 -> 94.3 | Improved |
| add global graph branch (GCN) to local subgraph model (ablation) | HC2-GNN · 20NG-L | accuracy (%): 90.7 -> 94.3 | Improved |
| use GRU-GGNN instead of GCN in subgraphs (ablation) | HC2-local · 20NG | accuracy (%): 85.7 -> 86.8 | Improved |
| use C2GC instead of rigid conductance partitioning (ablation) | HC2-GNN · 20NG | accuracy (%): 86.2 -> 89.6 | Improved |
| rigid conductance partitioning instead of GCN-only partitioning (ablation) | HC2-Simp · 20NG | accuracy (%): 85.7 -> 86.2 | No clear effect |
| apply HC2-GNN on short-text dataset | HC2-GNN vs TextING · MR | accuracy (%): 79.82 -> 80.22 | No clear effect |
| HC2-GNN vs TextING on long texts | HC2-GNN · 20NG | accuracy (%): 85.47 -> 89.66 | Improved |
| HC2-GNN-GT vs MSABertGCN | HC2-GNN-GT · R52 (transformer-enhanced) | accuracy (%): 96.68 -> 96.40 | No clear effect |
Closest papers by meaning
- TGLsta: Low-resource Textual Graph Learning with Semantic and Topological Awareness via LLMs (AAAI 2025 · similarity 0.87)
- Improved Graph Contrastive Learning for Short Text Classification (AAAI 2024 · similarity 0.86)
- Graph Classification via Reference Distribution Learning: Theory and Practice (NeurIPS 2024 · similarity 0.86)
- Label-free Node Classification on Graphs with Large Language Models (LLMs) (ICLR 2024 · similarity 0.86)
- Bootstrapping Heterogeneous Graph Representation Learning via Large Language Models: A Generalized Approach (AAAI 2025 · similarity 0.86)
- Translating Subgraphs to Nodes Makes Simple GNNs Strong and Efficient for Subgraph Representation Learning (ICML 2024 · similarity 0.86)
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.
