###
DOI:
工程科学与技术:2007,39(2):116-120
本文二维码信息
码上扫一扫!
基于多基准属性概化的知识归纳
(四川大学 计算机学院,四川 成都610065)
Knowledge Induction Based on Generalization of Multi benchmark Attribute
摘要
图/表
参考文献
相似文献
附件
本文已被:浏览 1408次   下载 3
投稿时间:2006-08-31    
中文摘要: 在分析基本面向属性归纳算法不足的基础上,针对中药方剂数据挖掘的特征提出基于面向关联属性概化层次的优化技术及算法。根据中医理论为各关联的维度创建概化层次及概念树,利用关联属性阈值与多基准属性阈值的相关性以提高归纳概化的效率和发现知识的准确率。实现面向关联属性归纳的中药方剂知识发现分析系统。通过实验验证了新算法的有效性,在相同条件下较传统算法的效率提高了27%以上。新系统能更快发现符合中医理论的领域知识和规则。
Abstract:Through analyzing the disfigurement of basic attribute oriented induction arithmetic, a new optimized technic and arithmetic based on generalizing hiberarchy of relevancy attribute-oriented was proposed for Chinese medicine prescription data mining. The efficiency of induction and the accuracy of discovery knowledge by the pertinency of relevancy-attribute-threshold and benchmark-attribute-threshold were enhanced,and generalizing hiberarchy and trees for relevancy attribute based on Chinese medical knowledge were established. A new system about Chinese medicine prescription knowledge discovery and analysis was implemented.The effectivity of the new arithmetic was proved by four extensive experiments. The efficiency was heighten above 27% comparing with the conventional arithmetic in the same condition. The speciality knowledge and rules in accordance with Chinese medicine theory can be found more rapidly by the new system.
文章编号:20070222     中图分类号:    文献标志码:
基金项目:国家自然科学基金(60473071;90409007);四川省科技攻关项目(2006Z01-027);科技部十一五支撑计划资助项目(2006BAI05A001)
作者简介:
引用文本:
刘齐宏,唐常杰,李川.基于多基准属性概化的知识归纳[J].工程科学与技术,2007,39(2):116-120.
.Knowledge Induction Based on Generalization of Multi benchmark Attribute[J].Advanced Engineering Sciences,2007,39(2):116-120.