BERT, a pre-trained Transformer model, has achieved ground-breaking performance on multiple NLP tasks. In this paper, we describe BERTSUM, a simple variant of BERT, for extractive summarization. Our system is the state of the art on the CNN/Dailymail dataset, outperforming the previous best-performed system by 1.65 on ROUGE-L. The codes to reproduce our results are available at https://github.com/nlpyang/BertSum
Fine-tune BERT for Extractive Summarization
BERTSUM, a variant of BERT, achieves top performance on extractive summarization using the CNN/Dailymail dataset with improvements in ROUGE-L.
- Year
- 2019
- Venue
- fine-tune-bert-for-extractive-summarization-1
- Authors
- 1
- Hosting
- Abstract onlyARXIV-DEFAULT
Cite
Notes
Only stored in your browser.
Attribution
- Abstract & full text
- arxiv.org/abs/1903.10318v2ARXIV-DEFAULT
- TL;DR
- Semantic Scholar