CSMED

CSMED是一个整合了九个公开发布的数据集的元数据集，提供了对医学和计算机科学领域325个系统文献综述的统一访问。该数据集旨在作为训练和评估自动化引文筛选模型的全面资源。此外，CSMED还包含一个专门设计用于评估全文出版物筛选任务的新数据集CSMED-FT。CSMED通过数据协调，解决了缺乏规范分割、适用性有限和数据集重叠的问题，为系统文献综述自动化领域的发展提供了重要支持。CSMED is a meta-dataset integrating nine publicly released datasets, offering unified access to 325 systematic literature reviews spanning the fields of medicine and computer science. It is designed as a comprehensive resource for training and evaluating automated citation screening models. Furthermore, CSMED contains a novel dataset, CSMED-FT, specifically developed to evaluate full-text publication screening tasks. Through data curation, CSMED addresses the issues of non-standardized splits, limited applicability, and dataset overlap, providing substantial support for the advancement of the automated systematic literature review field.