Hypergraph Benchmarks
UAT
Structure
1Graphs
1,190Nodes
4Classes
Coverage
78Results
1Variants
7Papers
- Attention Beyond Neighborhoods: Reviving Transformer for Graph Clustering Attention Beyond Neighborhoods: Reviving Transformer for Graph Clustering
- Aggregation-aware MLP: An Unsupervised Approach for Graph Message-passing Aggregation-aware MLP: An Unsupervised Approach for Graph Message-passing
- Non-Dissipative Graph Propagation for Non-Local Community Detection Non-Dissipative Graph Propagation for Non-Local Community Detection
- SynC: Synergistic Boosting of Structure and Representation for Deep Graph Clustering SynC: Synergistic Boosting of Structure and Representation for Deep Graph Clustering
- Provable Filter for Real-world Graph Clustering Provable Filter for Real-world Graph Clustering
- GraphLearner: Graph Node Clustering with Fully Learnable Augmentation GraphLearner: Graph Node Clustering with Fully Learnable Augmentation
- Simple Contrastive Graph Clustering Simple Contrastive Graph Clustering
4-class node classification US Airport Transactions hypergraph node classification (4 region classes, 1190 nodes). Uses node-hyperedge incidence and the released node features. Clique expansion changes the input by discarding hyperedge identity. Evaluated by F1.
Dataset details
Task
US Airport Transactions hypergraph node classification (4 region classes, 1190 nodes). The primary catalog metric is F1.
Data
US Airport Transactions hypergraph node classification (4 region classes, 1190 nodes). The cataloged artifact reports 1 graphs, 4 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 F1 (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 | F1 ↑ | ACC | Conductance | NMI | Paper |
|---|---|---|---|---|---|---|
| 1 | SCAGC | 0.7277
±
0.0016 | — | — | — | |
| 2 | SCGC | 0.7222
±
0.0097 | — | — | — | |
| 3 | AGE | 0.7174
±
0.0093 | — | — | — | |
| 4 | DAEGC | 0.6987
±
0.0054 | — | — | — | |
| 5 | GAE | 0.6808
±
0.0176 | — | — | — | |
| 6 | DFCN | 0.6718
±
0.0013 | — | — | — | |
| 7 | ARGA | 0.6430
±
0.0195 | — | — | — | |
| 8 | SynC KMeans | 0.5891
±
0.0102 | — | — | — | |
| 9 | SynC SSL | 0.5734
±
0.0023 | — | — | — | |
| 10 | SCGC | 0.5552
±
0.0087 | — | — | — | |
| 11 | NeuCGC | 0.5535
±
0.0126 | — | — | — | |
| 12 | HomoCAGC | 0.5531
±
0.0087 | — | — | — | |
| 13 | CCGC | 0.5524
±
0.0169 | — | — | — | |
| 14 | Vanilla GAE | 0.5439
±
0.0089 | — | — | — | |
| 15 | HSAN | 0.5420
±
0.0184 | — | — | — | |
| 16 | DAEGC | 0.5400
±
0.0100 Conductance
1.0000
±
0.0000 NMI
0.1600
±
0.0400 | — | 1.0000
±
0.0000 | 0.1600
±
0.0400 | |
| 17 | uAGNN | 0.5300
±
0.0200 Conductance
0.6900
±
0.2300 NMI
0.2900
±
0.0200 | — | 0.6900
±
0.2300 | 0.2900
±
0.0200 | |
| 18 | SSGC | 0.5279
±
0.0001 | — | — | — | |
| 19 | GraphLearner | 0.5277
±
0.0261 | — | — | — | |
| 20 | DGI | 0.5200
±
0.0300 Conductance
0.7000
±
0.1500 NMI
0.2400
±
0.0300 | — | 0.7000
±
0.1500 | 0.2400
±
0.0300 | |
| 21 | GRACE | 0.5200
±
0.0300 Conductance
0.9400
±
0.1900 NMI
0.2700
±
0.0100 | — | 0.9400
±
0.1900 | 0.2700
±
0.0100 | |
| 22 | AGC-DRR | 0.5197
±
0.0210 | — | — | — | |
| 23 | NS4GC | 0.5176
±
0.0052 | — | — | — | |
| 24 | SDCN | 0.5066
±
0.0149 | — | — | — | |
| 25 | DEAGC | 0.5033
±
0.0064 | — | — | — | |
| 26 | DMoN | 0.4900
±
0.0400 Conductance
0.6700
±
0.4400 NMI
0.2000
±
0.0400 | — | 0.6700
±
0.4400 | 0.2000
±
0.0400 | |
| 27 | MVGRL | 0.4900
±
0.0400 Conductance
0.9000
±
0.0300 NMI
0.2900
±
0.0200 | — | 0.9000
±
0.0300 | 0.2900
±
0.0200 | |
| 28 | BGRL | 0.4800
±
0.0200 Conductance
0.8800
±
0.1100 NMI
0.2400
±
0.0000 | — | 0.8800
±
0.1100 | 0.2400
±
0.0000 | |
| 29 | AE | 0.4787
±
0.0020 | — | — | — | |
| 30 | DEC | 0.4671
±
0.0012 | — | — | — | |
| 31 | SUBLIME | 0.4619
±
0.0087 | — | — | — | |
| 32 | DCRN | 0.4481
±
0.0087 | — | — | — | |
| 33 | VGAER | 0.4400
±
0.0300 Conductance
0.7500
±
0.1200 NMI
0.2000
±
0.0200 | — | 0.7500
±
0.1200 | 0.2000
±
0.0200 | |
| 34 | AFGRL | 0.3652
±
0.0089 | — | — | — | |
| 35 | GCA | 0.3572
±
0.0028 | — | — | — | |
| 36 | AGC-DRR | 0.3518
±
0.0032 | — | — | — | |
| 37 | AutoSSL | 0.3494
±
0.0087 | — | — | — | |
| 38 | MVGRL | 0.3288
±
0.0550 | — | — | — | |
| 39 | DFCN | 0.2579
±
0.0029 | — | — | — | |
| 40 | K-means | 0.2396
±
0.0051 | — | — | — | |
| 41 | SUBLIME | 0.1700
±
0.1300 Conductance
0.5000
±
0.2000 NMI
0.2000
±
0.0500 | — | 0.5000
±
0.2000 | 0.2000
±
0.0500 | |
| 42 | DFCN | — ACC
0.3361 NMI
0.2649 | 0.3361 | — | 0.2649 | |
| 43 | SSGC | — ACC
0.3674 NMI
0.0804 | 0.3674 | — | 0.0804 | |
| 44 | MVGRL | — ACC
0.4416 NMI
0.2153 | 0.4416 | — | 0.2153 | |
| 45 | SDCN | — ACC
0.5225 NMI
0.2161 | 0.5225 | — | 0.2161 | |
| 46 | MCGC | — ACC
0.4193 NMI
0.1664 | 0.4193 | — | 0.1664 | |
| 47 | FGC | — ACC
0.5303 NMI
0.2706 | 0.5303 | — | 0.2706 | |
| 48 | SCGC | — ACC
0.5658 NMI
0.2807 | 0.5658 | — | 0.2807 | |
| 49 | CCGC | — ACC
0.5634 NMI
0.2815 | 0.5634 | — | 0.2815 | |
| 50 | CGC | — ACC
0.4958 NMI
0.1749 | 0.4958 | — | 0.1749 |
SOTA performance
set a new SOTA at the time
matched the SOTA at the time