摘要
对医院特定患者的信息资源进行快速检索,能够有效提高医务人员的工作效率。对特定患者信息资源的检索,需要调整对信息特征项权重值,计算特征项向量梯度,完成患者信息的检索。传统方法捕获信息资源词项间的依赖关系,查询关联度,但忽略了特征项权重值的调整,导致检索精度低。提出基于动态自适应特征权重计算的信息资源快速检索方法。从医院患者信息资源库中提取目标信息与查询关键词间的语义路径,计算各语义路径权重,依据权重计算特定患者信息与各关键词间语义相关性。考虑信息特征项出现频率及所属文档在训练集中数量,对特征项权重值进行调整,通过考察特征项的分散度和向量梯度差,完成医院特定患者信息资源的快速检索。实验结果表明,该方法有效提高了特定患者信息检索精度,缩短了检索响应时间。
A quick retrieval method of information resources based on weight calculation of dynamic self - adaptive feature is proposed. The semantic path between target information and query keywords is extracted from information repository of hospital patients, and the weight of each semantic path is calculated. The semantic correlation between specific patient information and keywords is calculated according to weights. By considering the frequency of characteristic item of information and the number of document belong to the training set, the weight of characteristic item is adjusted. Through investigating the divergence of characteristic item and the vector gradient difference, the quick retrieval of specific information resource of hospital patient is completed. Simulation results show that this method can improve the precision of information retrieval for specific patient, which shortens the retrieval response time.
出处
《计算机仿真》
北大核心
2017年第12期389-392,共4页
Computer Simulation
关键词
医院
特定患者
信息资源
快速检索
Hospital
Specific patient
Information resources
Quick retrieval
