Transformer-based highlights extraction from scientific papers

作者：

Highlights：

• We propose a novel Transformer-based Highlights Extractor (THExt, in short).

• We achieve performance superior to state-of-the-art highlights extraction methods.

• Section-level context encoding turns out to be very effective for sentence ranking.

摘要

•We propose a novel Transformer-based Highlights Extractor (THExt, in short).•We achieve performance superior to state-of-the-art highlights extraction methods.•Section-level context encoding turns out to be very effective for sentence ranking.

论文关键词：Highlights extraction,Transformer model,Extractive summarization

论文评审过程：Received 28 December 2021, Revised 3 July 2022, Accepted 4 July 2022, Available online 8 July 2022, Version of Record 13 July 2022.

论文官网地址：https://doi.org/10.1016/j.knosys.2022.109382