sciencialab/grobid-evaluation
收藏Hugging Face2025-09-11 更新2025-11-01 收录
下载链接：
https://hf-mirror.com/datasets/sciencialab/grobid-evaluation
下载链接
链接失效反馈官方服务：
资源简介：
Grobid端到端评估数据集是一个用于测试GROBID软件性能的数据集，包括元数据提取、参考文献提取与解析、引用上下文识别和全文结构化等方面。该数据集由四个子集组成：PMC_sample_1943（来自PubMedCentral的1943篇文章）、biorxiv-10k-test-2000（来自bioRxiv的2000篇预印本文章）、PLOS_1000（来自PLOS的1000篇文章）和eLife_984（来自eLife的984篇文章）。每个子集都包含文章的PDF文件和相关XML文件。
The Grobid End 2 End evaluation dataset is designed for benchmarking the performance of the GROBID software, covering metadata extraction, bibliographical reference extraction and parsing, citation context identification, and full text structuring. The dataset consists of four subsets: PMC_sample_1943 (1943 articles from PubMedCentral), biorxiv-10k-test-2000 (2000 preprint articles from bioRxiv), PLOS_1000 (1000 articles from PLOS), and eLife_984 (984 articles from eLife). Each subset includes the PDF files of the articles and corresponding XML files.
提供机构：
sciencialab
