Hypergraph Benchmarks
House
Structure
1Graphs
1,290Nodes
340Edges
2Classes
100Node Feature Dim
Coverage
44Results
1Variants
4Papers
- Heterophily-Agnostic Hypergraph Neural Networks with Riemannian Local Exchanger Heterophily-Agnostic Hypergraph Neural Networks with Riemannian Local Exchanger
- Contrastive Learning for Non-Local Graphs with Multi-Resolution Structural Views Contrastive Learning for Non-Local Graphs with Multi-Resolution Structural Views
- ICLR 2022: Graph Auto-Encoder Via Neighborhood Wasserstein Reconstruction Graph Auto-Encoder Via Neighborhood Wasserstein Reconstruction ICLR 2022
- ICLR 2021: Hypergraph Benchmarks Hypergraph Benchmarks ICLR 2021
2-class node classification US House of Representatives voting hypergraph node classification (2 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
US House of Representatives voting hypergraph node classification (2 classes). The primary catalog metric is Accuracy.
Data
US House of Representatives voting hypergraph node classification (2 classes). The cataloged artifact reports 1 graphs, 2 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 ↑ | Completeness | Homogeneity | Silhouette | Paper |
|---|---|---|---|---|---|---|
| 1 | HealHGNN | 0.7718
±
0.0235 | — | — | — | |
| 2 | Deep-HGNN | 0.7526
±
0.0176 | — | — | — | |
| 3 | mechanism (shared) | 0.7470
±
0.0756 | — | — | — | |
| 4 | KHGNN | 0.7431
±
0.0287 | — | — | — | |
| 5 | SheafHyperGNN | 0.7384
±
0.0230 | — | — | — | |
| 6 | FrameHGNN | 0.7282
±
0.0222 | — | — | — | |
| 7 | ED-HNN | 0.7245
±
0.0228 | — | — | — | |
| 8 | HAN (full batch) | 0.7105
±
0.0226 | — | — | — | |
| 9 | PhenomNN | 0.7071
±
0.0235 | — | — | — | |
| 10 | AllSetTransformer | 0.6933
±
0.0220 | — | — | — | |
| 11 | CEGAT | 0.6909
±
0.0300 | — | — | — | |
| 12 | MLP | 0.6793
±
0.0233 | — | — | — | |
| 13 | AllDeepSets | 0.6782
±
0.0240 | — | — | — | |
| 14 | HNHN | 0.6780
±
0.0259 | — | — | — | |
| 15 | UniGCNII | 0.6725
±
0.0257 | — | — | — | |
| 16 | CECGN | 0.6280
±
0.0261 | — | — | — | |
| 17 | HGNN | 0.6139
±
0.0296 | — | — | — | |
| 18 | HCHA | 0.6136
±
0.0253 | — | — | — | |
| 19 | HyperGCN | 0.6013
±
0.0176 | — | — | — | |
| 20 | NWR-GAE | 0.5864
±
0.0561 | — | — | — | |
| 21 | MVGRL | 0.5351
±
0.0326 | — | — | — | |
| 22 | VGAE | 0.4932
±
0.0919 | — | — | — | |
| 23 | HyperGCN | 0.4831
±
0.0293 | — | — | — | |
| 24 | DGI | 0.4648
±
0.0797 | — | — | — | |
| 25 | GraphWave | 0.4596
±
0.0220 | — | — | — | |
| 26 | GAE | 0.4540
±
0.0999 | — | — | — | |
| 27 | GraphCL | 0.4486
±
0.0373 | — | — | — | |
| 28 | DeepWalk | 0.4121
±
0.0340 | — | — | — | |
| 29 | ARGVA | 0.4108
±
0.0485 | — | — | — | |
| 30 | Node2Vec | 0.4054
±
0.0162 | — | — | — | |
| 31 | RolX | 0.2567
±
0.1178 | — | — | — | |
| 32 | struc2vec | 0.2372
±
0.1369 | — | — | — | |
| 33 | DeepWalk | — Completeness
0.0100 Homogeneity
0.0100 Silhouette
0.2900 | 0.0100 | 0.0100 | 0.2900 | |
| 34 | Node2Vec | — Completeness
0.0100 Homogeneity
0.0100 Silhouette
0.3300 | 0.0100 | 0.0100 | 0.3300 | |
| 35 | RolX | — Completeness
1.0000 Homogeneity
1.0000 Silhouette
0.9900 | 1.0000 | 1.0000 | 0.9900 | |
| 36 | struc2vec | — Completeness
0.9900 Homogeneity
0.9900 Silhouette
0.4500 | 0.9900 | 0.9900 | 0.4500 | |
| 37 | GraphWave | — Completeness
1.0000 Homogeneity
1.0000 Silhouette
0.9900 | 1.0000 | 1.0000 | 0.9900 | |
| 38 | GAE | — Completeness
1.0000 Homogeneity
1.0000 Silhouette
0.9900 | 1.0000 | 1.0000 | 0.9900 | |
| 39 | VGAE | — Completeness
0.2700 Homogeneity
0.2500 Silhouette
0.2100 | 0.2700 | 0.2500 | 0.2100 | |
| 40 | ARGVA | — Completeness
0.2800 Homogeneity
0.2800 Silhouette
0.1900 | 0.2800 | 0.2800 | 0.1900 | |
| 41 | DGI | — Completeness
1.0000 Homogeneity
1.0000 Silhouette
0.9900 | 1.0000 | 1.0000 | 0.9900 | |
| 42 | GraphCL | — Completeness
1.0000 Homogeneity
1.0000 Silhouette
0.9900 | 1.0000 | 1.0000 | 0.9900 | |
| 43 | MVGRL | — Completeness
1.0000 Homogeneity
1.0000 Silhouette
0.9900 | 1.0000 | 1.0000 | 0.9900 | |
| 44 | NWR-GAE | — Completeness
1.0000 Homogeneity
1.0000 Silhouette
0.9900 | 1.0000 | 1.0000 | 0.9900 |
SOTA performance
set a new SOTA at the time
matched the SOTA at the time