Aspect Sentiment Triplet Extraction On Aste
Metriken
F1
Ergebnisse
Leistungsergebnisse verschiedener Modelle zu diesem Benchmark
Vergleichstabelle
| Modellname | F1 |
|---|---|
| grid-tagging-scheme-for-aspect-oriented-fine | 68.17 |
| contraste-supervised-contrastive-pre-training | 74.1 |
| aspect-sentiment-triplet-extraction-using | 69.61 |
| minicongts-a-near-ultimate-minimalist | 75.59 |
| span-level-bidirectional-cross-attention | 74.34 |
| towards-generative-aspect-based-sentiment | 72.16 |
| position-aware-tagging-for-aspect-sentiment | 63.83 |
| learning-span-level-interactions-for-aspect | 71.85 |
| first-target-and-opinion-then-polarity | 68.16 |
| a-unified-generative-framework-for-aspect | 67.62 |
| nonautoregressive-encoder-decoder-neural | 72.84 |