机械英文翻译Word下载.docx

机械英文翻译Word下载.docx

《机械英文翻译Word下载.docx》由会员分享，可在线阅读，更多相关《机械英文翻译Word下载.docx（13页珍藏版）》请在冰豆网上搜索。

Howcantheinformationisnotsubmerged,butfromtimetodiscoverusefulknowledgetoimproveinformationutilization."

Facethischallenge,dataminingcameintobeing,anddeveloprapidly,showingastrongvitality.

آآآآDataMiningistheso-calleddataminingfromalargenumberofincomplete,noisy,fuzzy,randomextractionoftherawdataimplicitinthemknowninadvance,butispotentiallyusefulinformationandknowledgeoftheprocessThebirthofdataminingtechnologyforpeopleonthedatabasetheresultsoflong-termresearchanddevelopment,anddataminingtechnologyFazhanthesametimeitinturnledintoadatabasetechnologymoreadvancedstage:

thedataHuanJingChuanTongbasicallydataoperationXing'

straditionalinformationsystemisonlyresponsiblefordata,deleteandmodifyoperationsinthedatabasecanberealizedonthebasisoftheworkisOLTP（OnLineTransactionProcesson-linetransactionprocessing）.Nowthatthegrowingaccumulationofdata,peopleneedtoanalyzethetypeofthedataenvironment,andsoitwasderivedfromthedatawarehousedatabase,youcanachieveasabasisforOLAP（OnLineAnalysisProcessOnlineAnalyticalProcessing）:

Withthemassivedatacollectionmaybeenhancedcomputerprocessingtechnologyandadvanceddataminingalgorithmsproposed,dataminingtechnologycannotonlyquerythedataofthepastandthetraverse,butalsotoidentifythepotentialvalueoverthepastlinksbetweenthedataandtocertainforms,andthusgreattomeeturgentneedsforknowledge.

آآآآDataminingisbasedontheformationoftheoriginaldatasourceofknowledge,Itcanbestructureddatasuchasrelationaldatabase,itcanbesemi-structured,suchastext,graphics,images,data,orevendistributiononthedifferentconfigurationdata.Thisarticlewillfocusononeforsemi-structureddatamining-WEB-baseddatamining,introducesitsbasicconceptsandtechniquesfrequentlyusedinthefinalbriefexplanationoftheXMLapplicationinwhich.

1,basedonWEBkeyconceptsofdatamining

1WhatisWEB-baseddatamining

Therapiddevelopmentofthecurrentnetwork,allsitesabound.ButinanincreasinglycompetitiveInterneteconomy,onlytowincustomersinordertoultimatelygainacompetitiveadvantage.Asasiteadministratororowner,shouldknowthatusersdoonhisWebsite,toknowwhichpartofthesitelikemostusers,whichallowsuserstofeeltired,outofasecurityvulnerabilitywhere,whatkindofchangeshavebroughtsignificantcustomersatisfactionandimprovethecontrary,whatkindofchangestheuserandsolost."

Knowthyself"

to"

knowyourself."

TheWEB-baseddataminingtechnologyisabletomeetthoseneeds.

WEB-baseddataminingontheexactdefinition,sofarnotveryclearandauthoritativestatement.Abroadthat:

WEB-baseddatamining,istousedataminingtechniquestoautomaticallydocumentfromthenetworkandservicediscoveryandextractionofinformationintheprocess.InTaiwan,differentopinions,thereisconsideredtobealargenumberofknowndatasamplesonthebasisoftheinherentcharacteristicsofdataobjects,andasabasisforapurposeintheWEBintheinformationextractionprocess.Atthesametime,scholarswillbethenetworkenvironmentincludedinthenetworkinformationretrievaldataminingandwebcontentdevelopmentandsoon.Inshort,WEB-baseddatamining（WebMining）isfromtheWorldWideWeb（WorldWideWeb）onaccesstorawdatafromahiddentapthepotentialofavailableknowledgeandultimatelyusedincommercialoperationstomeettheneedsofmanagers.

2,WEB-baseddataminingclassification

AccordingtodifferentobjectsexcavatedWecanWEB-baseddataminingisdividedintothreecategories:

WEB-basedcontentmining（WebContentMining）

Themining-basedWEB（WebStructureMining）

WEB-baseduseofmining（WebUsageMining）

（1）WEB-basedcontentmining

Theso-calledWEB-basedcontentminingisactuallyadocumentfromtheWEBandthedescriptionoftheaccesstoknowledge,WEBDocumentsminingandconcept-basedindexorsearchforAgenttechnologyshouldalsobeattributedtosuchresources.ManytypesofWebinformationresources,thecurrentWWWinformationresourceshasbecomethesubjectofnetworkinformationresources,butinadditionalargenumberofpeopledirectlyfromthewebcrawling,indexing,queryservicestoachievetheresources,theconsiderablepartoftheinformationishiddeninthedata（Ifthequestionsraisedbytheuserdynamicallygeneratedresults,thereisdatainthedatabasesystem,orsomeprivatedata）cannotbeindexed,sotheycannotprovideeffectiveretrievalmethod,whichforcesustodigouttheseelements.Iftheformsfromtheperspectiveofinformationresources,WEBcontentistext,images,audio,video,metadatasuchasthecompositionofthevariousformsofdata,whichwerefertoWEB-basedcontentminingisalsoamultimediadataMining.

2,basedonthestructureoftheminingWEB

ThistypeofminingistheoverallstructurefromtheWorldWideWebandwebpagesfoundonthelinkbetweenknowledgeoftheprocess,itismainlythepotentialofthelinkstructureminingWEBmode.Thisideacomesfromcitationanalysis,thatis,byanalyzingawebpagelinkandthenumberoflinksandtheobjectwastoestablishthelinkstructureofitsownmodeofWEB.Thismodelcanbeusedforwebpageclassificationandcanthusberelatedtoandassociatedwithdifferentdegreesofsimilaritybetweenpagesofinformation.WEBstructuremininghelpsusersfindrelatedtopicsintheauthorityofthesite,andsearchresultsontherankingofnetworkresourcesisverysignificant.

3,basedontheuseofminingWEB

WEB-baseduseofmining,alsoknownasWEBlogmining（WebLogMining）.Andthefirsttwominingapproachtotheon-linedataminingoftheoriginalobject,useWEB-basedminingfaceisintheprocessofinteractiontheuserandthenetworktoextractdataoutofsecond-hand.Thesedatainclude:

webserveraccesslogs,proxyserver,logging,userregistrationinformation,andwhenusersvisittheWebsitebehaviorandaction,andsoon.WEBusageminingthisdata11recordstothelogfile,andthenaccumulatedinthelogfileminingtounderstandtheuser'

sWebbehaviordatawithmeaning.Theexamplebeforeusfallintothistype.

Excavatedfromfivetothreeformswerecomparedwiththespecificcontentofwhichwillbefurtherdescribedbelow.

WEB-basedcontentmining:

unstructuredsemi-structured\textdocumenthypertextdocuments\Bagofwordsn-gramswordorphraseintheconceptofrelationaldataentities\TFIDFandstatisticalmachinelearningvariants（includingnaturallanguageprocessing）\returnclassclustermodeltoexplorethetextextractionrulestoexploretheestablishmentofmodel.

Themining-basedWEB:

semi-structureddatabaseformofweblinkstructureof\super-textdocumentlinks\boundarysignsOEMrelationaldatagraphgraphic\ProprietaryAlgorithmforILP（revised）oftheassociationrules\explorehigh-frequencysub-structureexcavationsitesystemStructuralclassificationclustering.

WEB-basedmininguse:

interactiveforms\serverlogrecordslogrecordsbrowser\relationaltablegraphics\Proprietarystatisticalmachinelearningalgorithm（revised）associationrules\siteconstructionandmanagementofsalesimprovedtocreateausermode.

3,thecharacteristicsofdataminingbasedonWEB

（1）Whatisthesemi-structured

Theso-calledsemi-structuredasopposedtothepurposesofstructuredandunstructured.Wecallthetraditionaldatabasedatafullystructureddata,whiletherearestillsome,suchasabook,apicturesocompletelywithoutstructureunstructureddata.Semi-structuredissomewhereinbetween,withtheimplicitmodel,informationstructure,irregular,non-stricttypesofconstraintsandsoon.Semi-structureddatamodelhasthefollowingcharacteristics

Priordata,afterthemodel;

Semi-structureddatamodelisusedtodescribethedatastructureofinformation,ratherthanmandatoryconstraintdatastructure;

Semi-structureddatamodelnon-precise,itcanonlybedescribedaspartofthedatastructuremayalsobeundervariousstagesofdataprocessingperspectivevaries;

Semi-structureddatamodel,maybeverylargeevenmorethanthesizeofthesourcedata,andwillcontinuouslyupdateasthedataisintheprocessofdynamicchange.

（2）WEBcharacteristicsofthedata

Webdataonthemostimportantfeatureisthesemi-structured.However,dataontheWebandtraditionaldatainthedatabaseisdifferentfromtraditionaldatabaseshavesomedatamodel,candescribethemodeltospecificdataandspecificorganizationsinaccordancewiththelawofacertainconcentrationordistributionofstorage,structuralstrong;

theWeb,thedataisverycomplex,nospecificmodeltodescribethedataforeachsite,allindependentlydesignedandthedataitselfhasareadmeanddynamicvariability,andthereforethedataontheWebisnotastrongstructural.AtthesametimeWebpagesisadescriptionoflevels,asinglesiteisinaccordancewiththestructureoftheirarchitecture,whichhassomestructural.Therefore,webelievethat