Hypergraph Benchmarks
Cora-CA
Structure
1Graphs
2,708Nodes
1,072Edges
7Classes
1433Node Feature Dim
Coverage
26Results
1Variants
4Papers
- TopoBench: A Framework for Benchmarking Topological Deep Learning TopoBench: A Framework for Benchmarking Topological Deep Learning
- ICML 2023: Node Embedding from Neural Hamiltonian Orbits in Graph Neural Networks Node Embedding from Neural Hamiltonian Orbits in Graph Neural Networks ICML 2023
- NeurIPS 2022: Augmentations in Hypergraph Contrastive Learning: Fabricated and Generative Augmentations in Hypergraph Contrastive Learning: Fabricated and Generative NeurIPS 2022
- ICLR 2021: Hypergraph Benchmarks Hypergraph Benchmarks ICLR 2021
7-class node classification Cora co-authorship hypergraph node classification (7 subject classes). Uses node-hyperedge incidence and the released node features. Clique expansion changes the input by discarding hyperedge identity. Evaluated by Accuracy.
Dataset details
Task
Cora co-authorship hypergraph node classification (7 subject classes). The primary catalog metric is Accuracy.
Data
Cora co-authorship hypergraph node classification (7 subject classes). The cataloged artifact reports 1 graphs, 7 target classes or tasks.
Features
The release represents higher-order membership with a node-hyperedge incidence relation and supplies node features where available. Some source datasets have no measured features and use documented synthetic features; clique expansion is an alternative representation that discards hyperedge identity.
Splits and evaluation
The catalog records these protocols or variants: Standard split. Evaluation uses Accuracy (higher is better). Exact masks or folds must come from the cited release.
Quirks and pitfalls
Do not silently replace incidence propagation with clique expansion or merge distinct releases sharing a short name. Random synthetic features, split seeds, hyperedge multiplicity, and isolated-node handling can dominate results and must be reported.
Sources
SOTA Progression
new SOTA
matches SOTA at that time
other proposed result
other baseline result
Leaderboard
| # | Model | Accuracy ↑ | Paper |
|---|---|---|---|
| 1 | HAN (full batch) | 0.8404
±
0.0102 | |
| 2 | AllSetTransformer | 0.8363
±
0.0147 | |
| 3 | UniGCNII | 0.8360
±
0.0114 | |
| 4 | HGNN | 0.8264
±
0.0165 | |
| 5 | HCHA | 0.8255
±
0.0097 | |
| 6 | EDGNN | 0.8236
±
0.0072 | |
| 7 | GIL | 0.8204
±
0.0127 | |
| 8 | AllDeepSets | 0.8197
±
0.0150 | |
| 9 | UniGNN2 | 0.8171
±
0.0142 | |
| 10 | HyperGCN | 0.7948
±
0.0208 | |
| 11 | HGCN | 0.7872
±
0.0042 | |
| 12 | HNHN | 0.7719
±
0.0149 | |
| 13 | CECGN | 0.7705
±
0.0126 | |
| 14 | A6 | 0.7621
±
0.0126 | |
| 15 | CEGAT | 0.7616
±
0.0119 | |
| 16 | A2 | 0.7562
±
0.0142 | |
| 17 | A1 | 0.7545
±
0.0154 | |
| 18 | A3 | 0.7534
±
0.0140 | |
| 19 | A4 | 0.7529
±
0.0156 | |
| 20 | GCN | 0.7516
±
0.0065 | |
| 21 | MLP | 0.7431
±
0.0189 | |
| 22 | Con | 0.7322
±
0.0165 | |
| 23 | Self | 0.7274
±
0.0153 | |
| 24 | A5 | 0.7261
±
0.0186 | |
| 25 | SetGNN | 0.7221
±
0.0151 | |
| 26 | A0 | 0.7185
±
0.0162 |
SOTA performance
set a new SOTA at the time
matched the SOTA at the time