基于概念格模型的分布式关联规则挖掘研究

《基于概念格模型的分布式关联规则挖掘研究》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、分类号。TP391密纽单位代码：10359学号：031805280616冬肥工学犬警HefeiUniversitVofTechnoIo．qy，一一硕士学位论文MASTERD1SSERTATIoH论文题目：基于概念格模型的分布式关联规则挖掘研究学位类别：学历硕士学科专业：(工稻!《i隘蓖)计算机应用技术作者姓名：导师姓名：完成时间：徐勇—————————————————————————————_+——————_——————————*————————胡学钢教授2006年05月ResearchonDistributedAssociationRulesMiningBasedonConceptL

2、atticesAbstractDistributedassociationrulesmining，emergingforsovlingtheassociationrulesminingindistributedenvironment，isoneofthemostactivetopicintheKDDfield，whichissignificantintheoryandapplication．Conceptlatticerepresentsdataandknowledgebytherelationsbetweenttheextensionandintensionofconcept，and

3、bythegeneralization／specializationrelationshipbetweenconcepts，whichissuitableforthestudyofassociationrulesmining．Thematurityofframeworkoflattice，leadinguptotherapidlyincreaseofsizeoflatticealongwiththeincreaseofdatabase’ssize，impedesconceptlattice’stheapplicationinKDD．Basedonstudyingtheconceptla

4、tticeanditsextension，thedistributedminingofassociationrulesbasedonprunedconceptlatticeswasdiscussedinthisdissertation．Themainpointsofthisdissertationarelistedasfollowing：(1)Basedonstudyingtherepresentationofdataandknowledgebasedonconcept，theprunedconceptlatticemodel(PCL)isdiscussedindetail．Theco

5、rrespondingformationalgorithmofPCLisgiventoo．(2)ComparingtherepresentationoffrequentitemsetsbasedonprunedconceptlatticeandApriori—likemodel．Thetheoryanalysisshowstheformerismoreconciseandmoreexactinrepresentationofassociationinoriginaldataset．(3)Studytheassociationrulesminingfrommultipledatasour

6、cesbasedonprunedconceptlattices，andthecorrespondingalgorithmUMPLisprovided．Theexperimentalevidenceshowsthevalidityofthisalgorithm．(4)Theprototypesystemthatisusedtominingassociationrulesbasedonconceptlatticeisimplemented．Keywords：datamining，conceptlattice，associationrules，distributedmining插图清单图l一

7、1数据挖掘是知识发现过程中的一个重要步骤⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯2图卜2小波变换实现数据挖掘结构图⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯g图21例2．1的哈斯图⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯13图2—2上F文2-1对应概念格的哈斯图⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯．16圈23上下文22对应的概念格L22的哈斯图⋯．⋯⋯⋯⋯．⋯⋯⋯⋯⋯⋯⋯⋯19图2—4上r文2-3对应的概念格L2—3的哈斯图⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯．20图2—5上下文2—1对应的EC