Relation Extraction on SemEval 2017
45.37Hierarchical F1HGNet
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| HGNetLearning Paradigm=Supervised2026.03 | 45.37 | 50.64 | 47.03 | |
| HGNetLearning Paradigm=Supervised, Loss Configuration=w/o LDHL2026.03 | 42.16 | 49.42 | 45.05 | |
| HGNetLearning Paradigm=Supervised, Loss Configuration=w/o LCAF Loss2026.03 | 38.58 | 43.28 | 41.19 | |
| GATLearning Paradigm=Supervised2026.03 | 34.47 | 46.19 | 39.25 | |
| GCNLearning Paradigm=Supervised2026.03 | 34.08 | 45.92 | 38.96 | |
| HGERELearning Paradigm=Supervised2026.03 | 33.81 | 45.73 | 38.63 | |
| GATLearning Paradigm=Supervised, Loss Configuration=w/o LDHL2026.03 | 33.52 | 45.88 | 38.43 | |
| PL-MarkerLearning Paradigm=Supervised2026.03 | 32.96 | 43.4 | 37.19 | |
| GCNLearning Paradigm=Supervised, Loss Configuration=w/o LDHL2026.03 | 32.82 | 45.72 | 37.99 | |
| PURELearning Paradigm=Supervised2026.03 | 28.94 | 41.35 | 34.92 | |
| llama-3.3-70b-versatileLearning Paradigm=Zero-Shot2026.03 | 23.65 | 25.38 | 24.12 | |
| openai/gpt-oss-120bLearning Paradigm=Zero-Shot2026.03 | 23.59 | 24.16 | 23.88 | |
| qwen/qwen3-32bLearning Paradigm=Zero-Shot2026.03 | 20.92 | 21.38 | 21.09 | |
| GPT-3.5 TurboLearning Paradigm=Zero-Shot2026.03 | 16.3 | 17.13 | 16.74 | |
| llama-3.1-8b-instantLearning Paradigm=Zero-Shot2026.03 | 14.11 | 14.46 | 14.24 |