Data-to-text evaluation on SFHOT
0.32Spearman Correlation (Naturalness)UniEval
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| UniEvalMetric Category=Trained, Training Dataset=Summ2023.07 | 0.32 | — | 0.249 | — | — | |
| DecompEvalMetric Category=Untrained2023.07 | 0.316 | — | 0.302 | — | — | |
| DIFFSCORE-FT2026.05 | 0.309 | — | 0.29 | — | 0.292 | |
| UniEvalMetric Category=Trained, Training Dataset=Dial2023.07 | 0.291 | — | 0.196 | — | — | |
| BARTScoreMetric Category=Trained, Training Dataset=CNNDM2023.07 | 0.288 | — | 0.235 | — | — | |
| BARTScore2026.05 | 0.272 | — | 0.211 | — | 0.199 | |
| MoverScoreMetric Category=Untrained2023.07 | 0.242 | — | 0.172 | — | — | |
| G-Eval2026.05 | 0.21 | — | 0.322 | — | 0.28 | |
| BERTScore2026.05 | 0.204 | — | 0.183 | — | 0.104 | |
| ROUGE-1Metric Category=Untrained2023.07 | 0.196 | — | 0.118 | — | — | |
| ROUGE-LMetric Category=Untrained2023.07 | 0.186 | — | 0.11 | — | — | |
| DIFFSCORE2026.05 | 0.186 | — | 0.244 | — | 0.191 | |
| BERTScoreMetric Category=Untrained2023.07 | 0.178 | — | 0.135 | — | — | |
| BARTScoreMetric Category=Untrained2023.07 | 0.165 | — | 0.158 | — | — | |
| GPTScore2026.05 | 0.141 | — | 0.222 | — | 0.113 | |
| UniEval2026.05 | 0.138 | — | 0.279 | — | 0.121 | |
| MoverScore2026.05 | 0.136 | — | 0.149 | — | 0.056 | |
| CTRLEvalMetric Category=Untrained2023.07 | 0.121 | — | 0.158 | — | — | |
| ROUGE-12026.05 | 0.113 | — | 0.116 | — | 0.022 | |
| ROUGE-L2026.05 | 0.102 | — | 0.089 | — | 0.025 | |
| BLEU2026.05 | 0.091 | — | 0.003 | — | 0.097 | |
| METEOR2026.05 | 0.087 | — | 0.012 | — | 0.087 | |
| ROUGE-22026.05 | 0.086 | — | 0.08 | — | 0.018 | |
| AlignScore2026.05 | 0.061 | — | 0.12 | — | 0.014 | |
| QuestEval2026.05 | 0.035 | — | 0.12 | — | 0.025 | |
| BARTSCOREFine-tuning=none2021.06 | — | 0.158 | — | — | — | |
| BARTSCORE2024.02 | — | 0.095 | — | 0.133 | — | |
| BARTSCORE-CNNFine-tuning=CNN/DM2021.06 | — | 0.19 | — | — | — | |
| BARTSCORE-CNN-PARAFine-tuning=CNN/DM + Paraphrase2021.06 | — | 0.211 | — | — | — | |
| BARTSCORE-PROMPTFine-tuning=CNN/DM + Paraphrase, Prompt combination=true2021.06 | — | 0.235 | — | — | — | |
| BERTScore2021.06 | — | 0.135 | — | — | — | |
| BERTSCORE2024.02 | — | 0.087 | — | 0.103 | — | |
| CTC2024.02 | — | 0.16 | — | 0.181 | — | |
| DecompEvalType=Instruction-tuned PLM (Zero-shot)2023.07 | — | 0.309 | — | — | — | |
| GPT-4 EVAL2024.02 | — | 0.364 | — | 0.436 | — | |
| HD-EVAL-NNIterative alignment training data percentage=25%, Aggregator type=NN2024.02 | — | 0.42 | — | 0.494 | — | |
| HD-EVAL-NNIterative alignment training data percentage=50%, Aggregator type=NN2024.02 | — | 0.432 | — | 0.51 | — | |
| MoverScore2021.06 | — | 0.172 | — | — | — | |
| PRISM2021.06 | — | 0.196 | — | — | — | |
| PRISM2024.02 | — | 0.131 | — | 0.164 | — | |
| ROUGE-12021.06 | — | 0.118 | — | — | — | |
| ROUGE-12024.02 | — | 0.031 | — | 0.035 | — | |
| ROUGE-22021.06 | — | 0.088 | — | — | — | |
| ROUGE-22024.02 | — | 0.042 | — | 0.06 | — | |
| ROUGE-L2021.06 | — | 0.11 | — | — | — | |
| ROUGE-L2024.02 | — | 0.038 | — | 0.048 | — | |
| UNIEVAL2024.02 | — | 0.305 | — | 0.35 | — | |
| UniEval (Dial)Training Dataset=Topical-Chat2023.07 | — | 0.244 | — | — | — | |
| UniEval (Summ)Training Dataset=SummEval2023.07 | — | 0.285 | — | — | — |