Target-Aspect-Sentiment Detection on FlightABSA
73.36F1 ScoreLLM
Evaluation Results
| Method | Links | |
|---|---|---|
| LLMMode=Fine-tuned2026.05 | 73.36 | |
| LLM-MvP# Shots=1002026.05 | 70.11 | |
| Para.Mode=Fine-tuned2026.05 | 69.74 | |
| MvPMode=Fine-tuned2026.05 | 68.67 | |
| LLM-MvP# Shots=502026.05 | 68.36 | |
| Self-consistency# Shots=1002026.05 | 68.11 | |
| Single-order# Shots=1002026.05 | 67.59 | |
| Self-consistency# Shots=502026.05 | 67.13 | |
| Single-order# Shots=502026.05 | 66.38 | |
| Gemma-3-27B ICLScenario=50 annotated examples, # Train=502026.03 | 64.6 | |
| Reasoning# Shots=1002026.05 | 63.09 | |
| LA-ABSA w/ DLOScenario=50 annotated examples, # Train=full2026.03 | 62.57 | |
| LA-ABSA w/ ParaphraseScenario=50 annotated examples, # Train=full2026.03 | 62.47 | |
| LLM-MvP# Shots=102026.05 | 62.22 | |
| Reasoning# Shots=502026.05 | 61.43 | |
| LA-ABSA w/ DLOScenario=10 annotated examples, # Train=full2026.03 | 61.4 | |
| Gemma-3-27B ICLScenario=10 annotated examples, # Train=102026.03 | 60.36 | |
| LLM-MvP# Shots=Avg.2026.05 | 59.95 | |
| LA-ABSA w/ ParaphraseScenario=10 annotated examples, # Train=full2026.03 | 59.87 | |
| Reasoning# Shots=Avg.2026.05 | 59.37 | |
| Self-consistency# Shots=102026.05 | 59.29 | |
| Single-order# Shots=102026.05 | 59.27 | |
| Self-consistency# Shots=Avg.2026.05 | 58.96 | |
| Reasoning# Shots=102026.05 | 58.62 | |
| Single-order# Shots=Avg.2026.05 | 57.79 | |
| Reasoning# Shots=02026.05 | 54.34 | |
| LA-ABSA w/ ParaphraseScenario=0 annotated examples, # Train=full2026.03 | 52.1 | |
| Gemma-3-27B (0-shot)Scenario=0 annotated examples, # Train=02026.03 | 51.81 | |
| LA-ABSA w/ DLOScenario=0 annotated examples, # Train=full2026.03 | 51.75 | |
| Self-consistency# Shots=02026.05 | 41.31 | |
| LLM-MvP# Shots=02026.05 | 39.11 | |
| Single-order# Shots=02026.05 | 37.93 |