Text Classification on 20ng (test)
89.5Accuracy (Test)ROBERTAGCN
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| ROBERTAGCNBackbone=RoBERTa2021.05 | 89.5 | — | — | — | — | — | — | |
| BertGCNBackbone=BERT-base2021.05 | 89.3 | — | — | — | — | — | — | |
| SGCTime (seconds)=19.06 ± 0.15, K=22019.02 | 88.5 | — | — | — | — | — | — | |
| SGC2021.05 | 88.5 | — | — | — | — | — | — | |
| TextGSL2025.12 | 88.32 | — | — | — | — | — | — | |
| GCNTime (seconds)=1205.1 ± 144.52019.02 | 87.9 | — | — | — | — | — | — | |
| LDGCN2025.12 | 87.79 | — | — | — | — | — | — | |
| TensorGCN2025.12 | 87.74 | — | — | — | — | — | — | |
| BertGATBackbone=BERT-base2021.05 | 87.4 | — | — | — | — | — | — | |
| DHTG2025.12 | 87.13 | — | — | — | — | — | — | |
| GTG2025.12 | 86.96 | — | — | — | — | — | — | |
| HyperGAT2025.12 | 86.62 | — | — | — | — | — | — | |
| ROBERTAGATBackbone=RoBERTa2021.05 | 86.5 | — | — | — | — | — | — | |
| TextGCN2025.12 | 86.34 | — | — | — | — | — | — | |
| TextGCN2021.05 | 86.3 | — | — | — | — | — | — | |
| BERTModel variant=BERT-base2021.05 | 85.3 | — | — | — | — | — | — | |
| TextSSL2025.12 | 85.26 | — | — | — | — | — | — | |
| RoBERTa2021.05 | 83.8 | — | — | — | — | — | — | |
| Base2025.10 | — | 83.85 | 72.78 | 82.98 | 63.79 | 81.8 | 70.12 | |
| Base (Clean)trained_on=ground truth data, noise=none2025.10 | — | 88.14 | 88.14 | 88.14 | 88.14 | 88.14 | 88.14 | |
| Clear2025.10 | — | 84.58 | 72.89 | 83.28 | 64.87 | 83.58 | 70.44 | |
| Co-Teaching2025.10 | — | 85.35 | 72.85 | 83.98 | 63.94 | 82.16 | 71.97 | |
| Delora2025.10 | — | 88.51 | 82.16 | 88.4 | 79.6 | 86.87 | 80.9 | |
| LAFT2025.10 | — | 85.64 | 74.17 | 83.95 | 64.37 | 82.64 | 71.58 | |
| LLM-basemodel=GPT-4o2025.10 | — | 72.15 | 72.15 | 72.15 | 72.15 | 72.15 | 72.15 | |
| NoiseAL2025.10 | — | 85.95 | 77.11 | 85.89 | 75.79 | 84.62 | 75.69 | |
| SelfMix2025.10 | — | 80.87 | 78.99 | 78.19 | 65.52 | 77.68 | 70.54 | |
| SENT2025.10 | — | 84.17 | 73.97 | 83.8 | 64.4 | 82.44 | 71.56 |