基于哈希学习算法的数字化医院非结构化数据智能检索  

Intelligent Retrieval of Unstructured Data in Digital Hospitals Based on Hash Learning Algorithms

作　　者：陈庆华[1] CHEN Qinghua(Binhai People’s Hospital,Yancheng 224500,Jiangsu,China)

机构地区：[1]滨海县人民医院,江苏省盐城市224500

出　　处：《中国卫生信息管理杂志》2025年第1期136-142,共7页Chinese Journal of Health Informatics and Management

摘　　要：目的 为解决当前数字化医院非结构化数据管理过程中利用Mapreduce算法实现目标数据检索时面临的数据冲突碰撞问题,提出基于哈希学习算法的数字化医院非结构化数据智能检索方法。方法 以多层Transformer(变换网络)编码单元为核心搭建预训练网络模型,将海量非结构化数据输入其中进行学习,抽取数据实体知识,并通过深度卷积神经网络提取数据包含的特征;利用哈希学习函数处理非结构化数据特征,为其匹配二进制哈希码;最后,依托哈希学习算法构建非结构化数据智能检索模型,依次分析所有哈希码样本与输入信息之间的相似性,找到最相似数据样本,即可完成非结构化数据智能检索。结果 应用该方法对两个数据集展开检索,最终得到的检索结果曲线下面积分别达到了0.81和0.8。结论 该方法能够满足数字化医院非结构化数据智能检索的质量要求。Objective In order to solve the problem of data conflict and collision in the process of unstructured data management in digital hospital,an intelligent retrieval method of unstructured data in digital hospital based on Hash learning algorithm is proposed.Methods A pre training network model is built with the multi-layer transformer(transformation network)coding unit as the core.Massive unstructured data is input into it for learning,data entity knowledge is extracted,and the features contained in the data are extracted through the deep convolutional neural network.The hash learning function is used to process the characteristics of unstructured data and match the binary hash code.Finally,the intelligent retrieval model of unstructured data is constructed based on Hash learning algorithm.The similarity between all hash code samples and input information is analyzed in turn,and the most similar data samples are found to complete the intelligent retrieval of unstructured data.Results The method was applied to retrieve two datasets,and the area under the retrieval result curve reached 0.81 and 0.8 respectively.Conclusion This method can meet the quality requirements of intelligent retrieval of unstructured data in digital hospital.

关 键 词：哈希学习算法 非结构化数据 数字化医院 特征提取 数据检索 

分 类 号：R-39[医药卫生] R319