Aspect Based Sentiment Analysis Absa On Asqp
평가 지표
F1 (R15)
평가 결과
이 벤치마크에서 각 모델의 성능 결과
모델 이름 | F1 (R15) | Paper Title | Repository |
|---|---|---|---|
| ChatGPT (gpt-3.5-turbo, few-shot) | 34.27 | MvP: Multi-view Prompting Improves Aspect Sentiment Tuple Prediction | - |
| DLO | 48.18 | Improving Aspect Sentiment Quad Prediction via Template-Order Data Augmentation | - |
| MvP (multi-task) | 52.21 | MvP: Multi-view Prompting Improves Aspect Sentiment Tuple Prediction | - |
| LEGO-ABSA (multi-task) | 46.10 | LEGO-ABSA: A Prompt-based Task Assemblable Unified Generative Framework for Multi-task Aspect-based Sentiment Analysis | - |
| ChatGPT (gpt-3.5-turbo, zero-shot) | 22.87 | MvP: Multi-view Prompting Improves Aspect Sentiment Tuple Prediction | - |
| Paraphrase | 46.93 | Aspect Sentiment Quad Prediction as Paraphrase Generation | - |
| TAS-BRET | 34.78 | Aspect Sentiment Quad Prediction as Paraphrase Generation | - |
| GAS | 45.98 | Towards Generative Aspect-Based Sentiment Analysis | |
| AugABSA | 50.01 | Generative Data Augmentation for Aspect Sentiment Quad Prediction | |
| MvP | 51.04 | MvP: Multi-view Prompting Improves Aspect Sentiment Tuple Prediction | - |
0 of 10 row(s) selected.