文本挖掘在Web日志數(shù)據(jù)預(yù)處理中的應(yīng)用研究.pdf

1、侖肥工學(xué)大HefeiUniversityofTechno警Iogy碩士學(xué)位論文MASTERDISSERTATIoN③論文題目：文在望塑壅型16日至塑塑苧竺型十墮里!型究學(xué)位類別：——高蝰塑墅!墮讀噠蘭蘭墮學(xué)科專業(yè)：工程領(lǐng)域L——瑚機(jī)廑￡l拄盤作者姓名：武洪萍導(dǎo)師姓名：旦國疊塾撞完成時(shí)間：—————』塑型旺旦———一AppliedResearchonWebTextMininginWebLogDataPreprocessingAbstrae

2、tWiththerapiddevelopmentandpopularizationofInternetnetworkhasbecomeaneffectiveplatformforpeopletoexchangeandmanipulateinformationWeputincreasingexpectationonthenetworkservice，hopingtoobtainricherandmoresuitableWebservice

3、sinanallroundwayHoweverduetoitsinherentopenness，dynamismandisomerism，usersusuallyfinditdifficulttopickupusefulinformationfromWWWinanexactandefficientwayTherefore，itbecomesurgenttofindawaytogettheexactirifermation，toobtai

4、nthepotentialknowledgebesidestheinformation，aswellastoprovidepersonalizedservicesThisisalsothehotstudysnbjectinWebdataminingWebMiningcanbedividedintothreeparts：wcbContentMiningWebStructureMining，andWebUsageMiningAmongthe

5、m，theusageminingtakesthewidestresearcharea，inwhichthedatapreprocessingplaysanimportantroleThisdissertationemphasizesonthedatapreprocessingtechnologyinWebUsageMining，analyzingandstudyingthekeyproblemsandtechnologyinvolved

6、inthepreprocessingThemainworksinthisdissertationareasfollowing：1Thebasicframe。procedureandminingtechnologyaresummarizedThestudyontheprocess，keytechnologyandmethodsoftheweblogpreprocessingismade2Therelatedtheoryoftextmini

7、ngaswellasitsanalyzingtechnologyisdiscussedsystematicallyandtheprocessofWebdataminingisproposed3Thetextclusteringalgorithmtotheprocessoftransactionidentificationisintroduced，suitableimprovementhasbeenmadeinassociationwit

8、hthewebcontents，basedontheanalysisofthedisadvantagesofthetraditionalrecognitionThetextclusteringalgorithmismodifiedinaccordancewiththerequirementofthemodifiedtransactionrecognition4WeproposeamodelforWebjournaldatapreproc