Hybrid multi-document summarization using pre-trained language models

作者：

Highlights：

• Introducing a multi-document summarizer, called HMSumm, based on pre-trained methods.

• Employing both extractive and abstractive methods to generate the final summary.

• Comparing the performance with the state-of-the-art single/multi-document summarizers.

摘要

•Introducing a multi-document summarizer, called HMSumm, based on pre-trained methods.•Employing both extractive and abstractive methods to generate the final summary.•Comparing the performance with the state-of-the-art single/multi-document summarizers.

论文关键词：Pre-trained language models,Extractive summarization,Abstractive summarization,Determinantal point process,Deep submodular network

论文评审过程：Received 18 September 2020, Revised 13 November 2021, Accepted 23 November 2021, Available online 18 December 2021, Version of Record 22 December 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.116292