Target Identification on DGTA
85.09BERTScoreQwen2.5-7B
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Qwen2.5-7BFine-tuning Strategy=Integrated fine-tuning2026.01 | 85.09 | 31.12 | 67.14 | 94.16 | 66.58 | |
| DeepSeek-R1-QwenFine-tuning Strategy=Integrated fine-tuning2026.01 | 84.94 | 30.96 | 66.99 | 94.62 | 66.76 | |
| Qwen2.5-7BFine-tuning Strategy=Two-stage fine-tuning, Evaluation Stage=Target identification stage2026.01 | 84.69 | 31.64 | 66.17 | 95.19 | 66.99 | |
| mT5Fine-tuning Strategy=Two-stage fine-tuning, Evaluation Stage=Target identification stage2026.01 | 84.29 | 28.82 | 65.71 | 86.59 | 60.16 | |
| RoBERTa-CRFFine-tuning Strategy=Two-stage fine-tuning, Evaluation Stage=Target identification stage2026.01 | 84.07 | 30.43 | 66.18 | 88.97 | 62.07 | |
| mT5-e2eFine-tuning Strategy=Integrated fine-tuning2026.01 | 83.42 | 29.67 | 65.88 | 89.73 | 62.06 | |
| Qwen2.5-7BPrompt Strategy=Instruction prompting2026.01 | 82.47 | 28.26 | 63.69 | 91.16 | 61.87 | |
| GLM4-9BPrompt Strategy=Instruction prompting2026.01 | 77.98 | 24.16 | 51.99 | 94.14 | 58.38 | |
| Llama3-8BPrompt Strategy=Instruction prompting2026.01 | 77.69 | 27.94 | 56.98 | 85.9 | 54.63 | |
| DeepSeek-V3Prompt Strategy=Instruction prompting2026.01 | 76.87 | 25.65 | 50.38 | 92.05 | 56.45 | |
| GPT-4oPrompt Strategy=Instruction prompting2026.01 | 73.72 | 21.51 | 43.99 | 94.34 | 54.09 |