Extractive Summarization on CNN/DailyMail anonymized (test)
42.69ROUGE-1LSTM + PN + BERT + RL
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| LSTM + PN + BERT + RLEncoder=LSTM, Decoder=Pointer Network, Pre-trained model=BERT, Reinforcement Learning=true2019.07 | 42.69 | 19.6 | 38.85 | |
| LSTM + PN + BERTEncoder=LSTM, Decoder=Pointer Network, Pre-trained model=BERT2019.07 | 42.39 | 19.51 | 38.69 | |
| LSTM + PN + RLEncoder=LSTM, Decoder=Pointer Network, Reinforcement Learning=true2019.07 | 41.85 | 18.93 | 38.13 | |
| Jadhav and Rajan2019.07 | 41.6 | 18.3 | 37.7 | |
| Zhou et al.2019.07 | 41.59 | 19.01 | 37.98 | |
| LSTM + PNEncoder=LSTM, Decoder=Pointer Network2019.07 | 41.56 | 18.77 | 37.83 | |
| Dong et al.2019.07 | 41.5 | 18.7 | 37.6 | |
| Chen and Bansal2019.07 | 41.47 | 18.72 | 37.76 |