TL
TL;DR
BERTSUM, a simple variant of BERT, for extractive summarization, is described, which is the state of the art on the CNN/Dailymail dataset, outperforming the previous best-performed system by 1.65 on ROUGE-L.
Abstract
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 this https URL
References18 items
1
BanditSum: Extractive Summarization as a Contextual Bandit
2
Neural Latent Extractive Document Summarization
3
Neural Document Summarization by Jointly Learning to Score and Select Sentences
4
The Best of Both Worlds: Combining Recent Advances in Neural Machine Translation
5