MedEASi
收藏资源简介:
MedEASi是由斯坦福大学等研究机构创建的医学文本简化数据集,包含1979对专家和简单文本。数据集通过专家、非专家和AI的协作注释,支持四种文本转换:详述、替换、删除和插入,以实现文本的可控简化。数据集涵盖广泛的医学子领域,旨在提高医学文本的可理解性和可访问性,解决医学文本复杂性问题。
MedEASi is a medical text simplification dataset developed by Stanford University and other research institutions, which comprises 1979 pairs of expert medical texts and their simplified plain-language counterparts. The dataset is constructed through collaborative annotation involving medical domain experts, non-expert users and AI agents, and supports four core text transformation operations: elaboration, replacement, deletion and insertion, to enable controllable medical text simplification. Spanning a wide range of medical subfields, this dataset is designed to enhance the comprehensibility and accessibility of medical texts, and address the complexity challenges associated with medical textual materials.




