數(shù)據(jù)挖掘在數(shù)據(jù)庫入侵檢測中的應用研究.pdf

1、哈爾濱工程大學碩士學位論文數(shù)據(jù)挖掘在數(shù)據(jù)庫入侵檢測中的應用研究姓名：姜平平申請學位級別：碩士專業(yè)：計算機軟件與理論指導教師：劉大昕20061201哈爾濱丁程大學碩十學位論文AbstractDatebaseManagementSystem①BMS)isoneofthethreeplatformsofinformationsystemsSOitssafetyisanimportantcomponentoftheinformationsafe

2、tyDatabaseintrusiondetectionisthefmalbarrierthatensuresthedatabase’Ssafetybutnowtheresearchonitisstillinitsinfancymostdetectionmethodsarcbased011dataminingtechnologyTheauthorputsemphasesontheimprovementofApnorialgorithmw

3、hichisaclassicalalgorithmofminingassociationrules，andthenappliesittothedatabaseintrusiondetectionsystemInthelycoccssoftheApriorialgorithm，therearcamassofuselessitcmsetswhosesupportis0aregeneratedItiscausedbylinkingtwodat

4、asctswithoutconsideringthemutuallyexclusiveattributesAccordinglytheauthorpropoundstheconceptofmutuallyexclusiveattribute，anddividesitintotwokinds：theobviouslymutuallyexclusiveattributeandtheimplicitlymutuallyexclusiveatt

5、ributeTheimprovedprcpmcessingalgorithmisproposedfortheobviouslymutuallyexclusiveattributeDuringthetransformationofthedala慨theobviouslymutuallyexclusiveattributesaremarkedThelinkoftheobviouslymutuallyexclusiveattributesca

6、nbeavoidedaccordingtothemarksTheApriorialgorithmisimprovedtodealwiththeimplicitlymutuallyexclusiveattributeandtojudgcwhetherthetwoattributesalemutuallyexclusivebeforelinkillgtwoitemsetsByoverleapingconnectionoperationofm

7、utuallyexclusiveattributes，thenumberofk—itcmsctsisr既lucedSOthattimeofminingfiequentitemsetsissavedWiththetwoimprovedflgofithms，theauthordesignsaselfadaptivemodelofdatabaseintrusiondetectionsystemAccordingtothelhnitationo