NCBI Disease Corpus

“NCBI 疾病语料库在提及和概念级别进行了充分注释，可作为生物医学自然语言处理社区的研究资源。语料库特征 793 PubMed 摘要 6,892 疾病提及 790 个独特的疾病概念 医学主题词 (MeSH®) Online Mendelian Inheritance in Man (OMIM®) 91% 的提及映射到单个疾病概念，分为训练集、开发集和测试集 语料库注释 14 个注释器 每个文档两个注释器（随机配对） 三个注释阶段 检查整个语料库注释的一致性”The NCBI Disease Corpus is fully annotated at both the mention and concept levels, serving as a research resource for the biomedical natural language processing community. Corpus Characteristics: - 793 PubMed abstracts - 6,892 disease mentions - 790 unique disease concepts, with annotations referencing Medical Subject Headings (MeSH®) and Online Mendelian Inheritance in Man (OMIM®) - 91% of mentions are mapped to a single disease concept - The corpus is divided into training, development, and test sets Corpus Annotation: - 14 annotators in total - Two annotators were randomly paired for each document - Three annotation stages were conducted - Annotation consistency across the entire corpus was verified

搜集汇总

背景与挑战

背景概述

NCBI Disease Corpus是一个生物医学自然语言处理领域的研究资源，包含793篇PubMed摘要中的6,892个标准化疾病提及，采用多阶段严格标注流程。该数据集发布于2012年，提供训练/开发/测试集划分，主要用于疾病命名实体识别任务。