基于频繁概念格的电子病历关联规则挖掘研究  

Research of Association Rules Mining of Electronic Patient Record Based on Frequent Concept Lattice

作　　者：丁卫平[1] 施佺[1] 管致锦[1] 石振国[1] 

机构地区：[1]南通大学计算机科学与技术学院,江苏南通226019

出　　处：《微电子学与计算机》2008年第8期125-128,132,共5页Microelectronics & Computer

基　　金：江苏省高校自然科学基金项目(05KJB520107,07KJB520096);南通大学自然科学基金项目(05Z061);南通大学通信与信息系统学科科技创新项目

摘　　要：提出了一种基于频繁概念格的FC Lattice电子病历关联规则挖掘算法,该算法在渐进式建格的同时能发现最大频繁项集和提取相应关联规则,并进行基于FC Lattice算法的电子病历挖掘相关实验,结果表明该算法在运行速度和挖掘性能上都是高效的.The association roles mining algorithm （FC_ Lattice） of the electronic patient record based on the frequent concept lattice is put forward. And the algrithm can constuct the corresponding lattice incrementally. At the same time it also can find maximal frequent item-set and extract the corresponding association rules. Finally the mining experiment of the electronic patient record is designed. The results show that it is better on both sides of the running speed and the mining capability.

关 键 词：频繁项集 概念格 电子病历 关联规则 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]