LaySumm
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/WING-NUS/scisumm-corpus/blob/master/README_Laysumm.md
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含了572篇研究论文及其相应的黄金标准大众摘要，这些摘要以通俗易懂的语言概括了研究论文的核心内容。数据集中的摘要平均长度为84个词汇，涵盖了包括癫痫、考古和材料工程等不同领域的论文。数据规模方面，共有572篇训练论文和37篇盲测论文。该任务旨在进行大众化的摘要编写。
This dataset contains 572 research papers and their corresponding gold-standard lay summaries, which summarize the core content of each paper in plain, accessible language. The average length of these summaries is 84 words, and the dataset covers research papers across diverse fields including epilepsy, archaeology, and materials engineering. In terms of dataset size, there are 572 training papers and 37 blind-test papers. This task targets lay summarization.
提供机构：
Workshop on Scholarly Document Processing @EMNLP 2020
