Metric: F1 (R15) (higher is better)
| # | Model↕ | F1 (R15)▼ | Extra Data | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | MvP (multi-task) | 64.74 | Yes | MvP: Multi-view Prompting Improves Aspect Sentim... | 2023-05-22 | Code |
| 2 | MvP | 64.53 | No | MvP: Multi-view Prompting Improves Aspect Sentim... | 2023-05-22 | Code |
| 3 | Paraphrase | 63.06 | No | Aspect Sentiment Quad Prediction as Paraphrase G... | 2021-10-02 | Code |
| 4 | DLO | 62.95 | No | Improving Aspect Sentiment Quad Prediction via T... | 2022-10-19 | Code |
| 5 | LEGO-ABSA (multi-task) | 62.3 | Yes | - | - | - |
| 6 | Gemma-3-27B (50-shot, self-consistency learning) | 62.12 | No | Do we still need Human Annotators? Prompting Lar... | 2025-02-18 | Code |
| 7 | GAS | 60.63 | No | - | - | Code |
| 8 | TAS-BERT | 57.51 | No | Aspect Sentiment Quad Prediction as Paraphrase G... | 2021-10-02 | Code |
| 9 | Gemma-3-27B (10-shot, self-consistency learning) | 54.37 | No | Do we still need Human Annotators? Prompting Lar... | 2025-02-18 | Code |