semantic scholar ORC

我们介绍了S2ORC，这是一个涵盖许多学科的81.1m英语学术论文的大型语料库。语料库由丰富的元数据，论文摘要，已解析的书目参考以及8.1M开放获取论文的结构化全文组成。全文使用自动检测到的引文，图形和表格的内联引用进行注释，每个引用都链接到其相应的纸质对象。在S2ORC中，我们将来自数百家学术出版商和数字档案馆的论文汇总到一个统一的来源中，并创建迄今为止最大的可公开获取的机器可读学术文本集合。我们希望此资源将促进对学术文本进行文本挖掘的工具和任务的研究和开发。

We introduce S2ORC, a large corpus of 81.1 million English academic papers spanning numerous disciplines. This corpus provides rich metadata, paper abstracts, parsed bibliographic references, as well as structured full-texts for 8.1 million open access papers. The full texts are annotated with inline citations for automatically detected citations, figures and tables, with each citation linked to its corresponding paper object. In S2ORC, we aggregate papers from hundreds of academic publishers and digital archives into a unified resource, creating the largest publicly available machine-readable academic text collection to date. We anticipate that this resource will facilitate the research and development of tools and tasks for academic text mining.

搜集汇总

背景与挑战

背景概述

S2ORC是一个大型多学科学术论文语料库，包含81.1百万篇英语论文，其中810万篇为开放获取的结构化全文，并具有自动注释的引文、图形和表格链接。该数据集整合了数百家出版商的资源，旨在为学术文本挖掘提供机器可读的统一数据源，以促进相关工具和任务的研究开发。