EHR-RelB
收藏arXiv2020-10-30 更新2024-06-21 收录
下载链接：
https://github.com/babylonhealth/EHR-Rel
下载链接
链接失效反馈官方服务：
资源简介：
EHR-RelB数据集是由Babylon Health伦敦团队创建的，专注于电子健康记录（EHR）中的生物医学概念相关性分析。该数据集包含3630对概念，是现有数据集的六倍大小，通过自动从EHR中检索频繁共现的概念对来确保其相关性。数据集的创建旨在解决现有数据集规模小且手动选择概念对的问题，通过提供一个大规模、高质量的基准来测试和改进概念相关性模型。该数据集的应用领域包括医疗信息检索、临床决策支持等，旨在提高医疗服务的效率和质量。
The EHR-RelB dataset was created by the London team of Babylon Health, focusing on biomedical concept relevance analysis in electronic health records (EHRs). Comprising 3,630 concept pairs, the dataset is six times the size of existing datasets, with the relevance of these pairs ensured by automatically retrieving frequently co-occurring concept pairs from EHRs. The dataset was developed to address the limitations of small scale and manual concept pair selection in existing datasets, serving as a large-scale, high-quality benchmark for testing and improving concept relevance models. Its applicable domains include medical information retrieval, clinical decision support and other related fields, with the goal of enhancing the efficiency and quality of healthcare services.
提供机构：
Babylon Health 伦敦
创建时间：
2020-10-30
