CliCR

CliCR是一个专为医疗领域机器阅读理解设计的大型数据集，由比利时安特卫普大学的计算语言学与心理语言学研究中心创建。该数据集包含约100,000个关于临床病例报告的填空式查询，旨在通过这些查询评估机器在医疗文本理解上的表现。数据集内容丰富，平均每篇报告约1,500个词汇，涵盖多种医疗实体。创建过程中，研究团队从BMJ病例报告中提取病例描述，通过自动化和人工校验确保数据质量。CliCR的应用领域主要集中在临床决策支持，帮助医生从大量医疗文本中快速获取关键信息，提高诊疗效率。

CliCR is a large-scale dataset designed for machine reading comprehension in the medical domain, created by the Center for Computational Linguistics and Psycholinguistics at the University of Antwerp, Belgium. This dataset contains approximately 100,000 fill-in-the-blank queries focused on clinical case reports, intended to evaluate machine performance in medical text comprehension. The dataset is content-rich, with each report averaging around 1,500 words and covering a wide range of medical entities. During its development, the research team extracted case descriptions from BMJ case reports and ensured data quality through automated and manual verification. The primary application scenarios of CliCR center on clinical decision support, assisting clinicians in quickly obtaining key information from massive medical texts to improve diagnostic and treatment efficiency.