Abstractive Summarization on Stack ConvoSumm 1.0 (test)
39.73ROUGE-1BART-arg-graph
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| BART-arg-graphInput representation=arg-graph, Training points=2002021.06 | 39.73 | 11.17 | 35.52 | |
| BART-argPre-trained Model=BART-large, Training Samples=200 points, Input Format=Argument-mining input2021.06 | 39.73 | 11.17 | 35.52 | |
| BARTInput representation=vanilla, Training points=2002021.06 | 39.61 | 10.98 | 35.35 | |
| BARTPre-trained Model=BART-large, Training Samples=200 points, Input Format=Vanilla2021.06 | 39.61 | 10.98 | 35.35 | |
| BART-arg-filteredInput representation=arg-filtered, Training points=2002021.06 | 39.4 | 10.98 | 35.51 | |
| DIONYSUSModel Size=Large, Evaluation Protocol=Zero-shot, Strategy=Better ROUGE2022.12 | 28.5 | 5.6 | 17.6 | |
| PEGASUSModel Size=Large, Evaluation Protocol=Zero-shot2022.12 | 26.7 | 4.8 | 15.2 | |
| DIONYSUSModel Size=Large, Evaluation Protocol=Zero-shot, Strategy=All G2022.12 | 26.3 | 5.4 | 16.8 | |
| DIONYSUSModel Size=Large, Evaluation Protocol=Zero-shot, Strategy=All P2022.12 | 24.5 | 4.3 | 15 | |
| GSG+Evaluation Protocol=Zero-shot2022.12 | 21.2 | 3.5 | 15.1 | |
| T5v1.1Model Size=Large, Evaluation Protocol=Zero-shot2022.12 | 15.6 | 2.4 | 11 |