Sentiment Quadruplet Extraction on Rest SemEval 2016 (test)
57.92PrecisionDLO
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| DLOBackbone=T5-base, Decoding=Beam Search (k=5)2022.10 | 57.92 | 61.8 | 59.79 | |
| ILOBackbone=T5-base, Decoding=Beam Search (k=5)2022.10 | 57.58 | 61.17 | 59.32 | |
| ParaphraseModel Type=Generative2022.10 | 56.63 | 59.3 | 57.92 | |
| GASModel Type=Generative2022.10 | 54.54 | 57.62 | 56.04 | |
| TASO-BERT-LinearModel Type=Discriminative/Tagging2022.10 | 49.73 | 40.7 | 44.77 | |
| TASO-BERT-CRFModel Type=Discriminative/Tagging2022.10 | 48.65 | 39.68 | 43.71 | |
| Extract-Classify-ACOSModel Type=Two-stage2022.10 | 38.4 | 50.93 | 43.77 | |
| HGCN-BERT+BERT-TFMModel Type=Discriminative/Tagging2022.10 | 27.4 | 26.41 | 26.9 | |
| HGCN-BERT+BERT-LinearModel Type=Discriminative/Tagging2022.10 | 25.36 | 24.03 | 24.68 |