Target-Aspect-Sentiment Detection on Rest15
72.41F1 ScoreLLM
Evaluation results
| Method | Links | |
|---|---|---|
| LLMMode=Fine-tuned2026.05 | 72.41 | |
| Gemma 4Size=31B, Mode=Fine-tuned2026.05 | 72.41 | |
| LLM-MvPShots=1002026.05 | 70.5 | |
| LLM-MvP# Shots=1002026.05 | 70.5 | |
| LLM-MvPSetting=100-shot2026.05 | 70.5 | |
| Orca 2Size=13B, Mode=Fine-tuned2026.05 | 70.49 | |
| Orca 2Size=7B, Mode=Fine-tuned2026.05 | 69.74 | |
| LLM-MvP# Shots=502026.05 | 68.18 | |
| Self-consistency# Shots=1002026.05 | 65.94 | |
| E2TP2026.05 | 65.8 | |
| PGSO2026.05 | 65.4 | |
| Single-order# Shots=1002026.05 | 65.27 | |
| MvPMode=Fine-tuned2026.05 | 64.53 | |
| MvP2026.05 | 64.53 | |
| Self-consistency# Shots=502026.05 | 63.56 | |
| LLM-MvPShots=102026.05 | 63.29 | |
| LLM-MvP# Shots=102026.05 | 63.29 | |
| ParaphraseScenario=Full human set, # Train=full2026.03 | 63.06 | |
| Para.Mode=Fine-tuned2026.05 | 63.06 | |
| Paraphrase2026.05 | 63.06 | |
| Single-order# Shots=502026.05 | 63.05 | |
| DLOScenario=Full human set, # Train=full2026.03 | 62.95 | |
| Gemma-3-27B ICLScenario=50 annotated examples, # Train=502026.03 | 62.12 | |
| LLM-MvP# Shots=Avg.2026.05 | 60.71 | |
| GAS2026.05 | 60.63 | |
| Self-consistency# Shots=102026.05 | 59.43 | |
| Single-order# Shots=102026.05 | 58.82 | |
| LA-ABSA w/ DLOScenario=50 annotated examples, # Train=full2026.03 | 58.4 | |
| Self-consistency# Shots=Avg.2026.05 | 57.75 | |
| gpt-oss-120BShots=100, Prompting Strategy=CoT2026.05 | 57.08 | |
| Reasoning# Shots=1002026.05 | 57.08 | |
| Single-order# Shots=Avg.2026.05 | 56.91 | |
| LA-ABSA w/ ParaphraseScenario=50 annotated examples, # Train=full2026.03 | 56.21 | |
| Reasoning# Shots=502026.05 | 55.06 | |
| Gemma-3-27B ICLScenario=10 annotated examples, # Train=102026.03 | 54.47 | |
| Gemma-3-27BShots=10, Prompting Strategy=SC2026.05 | 54.47 | |
| Reasoning# Shots=Avg.2026.05 | 52.72 | |
| gpt-oss-120BShots=10, Prompting Strategy=CoT2026.05 | 50.55 | |
| Reasoning# Shots=102026.05 | 50.55 | |
| LA-ABSA w/ DLOScenario=10 annotated examples, # Train=full2026.03 | 49.23 | |
| LA-ABSA w/ ParaphraseScenario=10 annotated examples, # Train=full2026.03 | 49.09 | |
| gpt-oss-120BShots=0, Prompting Strategy=CoT2026.05 | 48.18 | |
| Reasoning# Shots=02026.05 | 48.18 | |
| ChatABSAShots=102026.05 | 45.93 | |
| QAIEScenario=50 annotated examples, # Train=141.82026.03 | 45.01 | |
| EDAScenario=50 annotated examples, # Train=5502026.03 | 44.14 | |
| Self-consistency# Shots=02026.05 | 42.08 | |
| LLM-MvPShots=02026.05 | 40.88 | |
| LLM-MvP# Shots=02026.05 | 40.88 | |
| Single-order# Shots=02026.05 | 40.48 | |
| DLOScenario=50 annotated examples, # Train=502026.03 | 39.54 | |
| ChatABSAShots=02026.05 | 39.21 | |
| DS²-ABSAScenario=50 annotated examples, # Train=21,7k2026.03 | 37.94 | |
| ParaphraseScenario=50 annotated examples, # Train=502026.03 | 36.92 | |
| Gemma-3-27B (0-shot)Scenario=0 annotated examples, # Train=02026.03 | 30.36 | |
| Gemma-3-27BShots=0, Prompting Strategy=SC2026.05 | 30.36 | |
| EDAScenario=10 annotated examples, # Train=1102026.03 | 30.18 | |
| DS²-ABSAScenario=10 annotated examples, # Train=21,1k2026.03 | 29.04 | |
| LA-ABSA w/ ParaphraseScenario=0 annotated examples, # Train=full2026.03 | 27.7 | |
| LA-ABSA w/ DLOScenario=0 annotated examples, # Train=full2026.03 | 25.82 | |
| QAIEScenario=10 annotated examples, # Train=26.42026.03 | 20.11 | |
| DLOScenario=10 annotated examples, # Train=102026.03 | 15.84 | |
| ParaphraseScenario=10 annotated examples, # Train=102026.03 | 8.75 |