數(shù)據(jù)挖掘外文翻譯參考文獻數(shù)據(jù)挖掘外文翻譯參考文獻(文檔含中英文對照即英文原文和中文翻譯)外文：WhatisDataMining?Simplystated,dataminingreferstoextractingor“mining”knowledgefromlargeamountsofdata.Thetermisactuallyamisnomer.Rememberthattheminingofgoldfromrocksorsandisreferredtoasgoldminingratherthanrockorsandmining.Thus,“datamining”shouldhavebeenmoreappropriatelynamed“knowledgeminingfromdata”,whichisunfortunatelysomewhatlong.“Knowledgemining”,ashorterterm,maynotreflecttheemphasisonminingfromlargeamountsofdata.Nevertheless,miningisavividtermcharacterizingtheprocessthatfindsasmallsetofpreciousnuggetsfromagreatdealofrawmaterial.Thus,suchamisnomerwhichcarriesboth“data”and“mining”becameapopularchoice.Therearemanyothertermscarryingasimilarorslightlydifferentmeaningtodatamining,suchasknowledgeminingfromdatabases,knowledgeextraction,data/patternanalysis,dataarchaeology,anddatadredging.Manypeopletreatdataminingasasynonymforanotherpopularlyusedterm,“KnowledgeDiscoveryinDatabases”,orKDD.Alternatively,othersviewdataminingassimplyanessentialstepintheprocessofknowledgediscoveryindatabases.Knowledgediscoveryconsistsofaniterativesequenceofthefollowingsteps:·datacleaning:toremovenoiseorirrelevantdata,·dataintegration:wheremultipledatasourcesmaybecombined,·dataselection:wheredatarelevanttotheanalysistaskareretrievedfromthedatabase,·datatransformation:wheredataaretransformedorconsolidatedintoformsappropriateforminingbyperformingsummaryoraggregationoperations,forinstance,·datamining:anessentialprocesswhereintelligentmethodsareappliedinordertoextractdatapatterns,·patternevaluation:toidentifythetrulyinterestingpatternsrepresentingknowledgebasedonsomeinterestingnessmeasures,and·knowledgepresentation:wherevisualizationandknowledgerepresentationtechniquesareusedtopresenttheminedknowledgetotheuser.Thedataminingstepmayinteractwiththeuseroraknowledgebase.Theinterestingpatternsarepresentedtotheuser,andmaybestoredasnewknowledgeintheknowledgebase.Notethataccordingtothisview,dataminingisonlyonestepintheentireprocess,albeitanessentialonesinceituncovershiddenpatternsforevaluation.Weagreethatdataminingisaknowledgediscoveryprocess.However,inindustry,inmedia,andinthedatabaseresearchmilieu,theterm“datamining”isbecomingmorepopularthanthelongertermof“knowledgediscoveryindatabases”.Therefore,inthisbook,wechoosetousetheterm“datamining”.Weadoptabroadviewofdataminingfunctionality:dataminingistheprocessofdiscoveringinterestingknowledgefromlargeamountsofdatastoredeitherindatabases,datawarehouses,orotherinformationrepositories.Basedonthisview,thearchitectureofatypicaldataminingsystemmayhavethefollowingmajorcomponents:1.Database,datawarehouse,orotherinformationrepository.Thisisoneorasetofdatabases,datawarehouses,spreadsheets,orotherkindsofinformationrepositories.Datacleaninganddataintegrationtechniquesmaybeperformedonthedata.2.Databaseordatawarehouseserver.Thedatabaseordatawarehouseserverisresponsibleforfetchingtherelevantdata,basedontheuser’sdataminingrequest.3.Knowledgebase.Thisisthedomainknowledgethatisusedtoguidethesearch,orevaluatetheinterestingnessofresultingpatterns.Suchknowledgecanincludeconcepthierarchies,usedtoorganizeattributesorattributevaluesintodifferentlevelsofabstraction.Knowledgesuchasuserbeliefs,whichcanbeusedtoassessapattern’sinterestingnessbasedonitsunexpectedness,mayalsobeincluded.Otherexamplesofdomainknowledgeareadditionalinterestingnessconstraintsorthresholds,andmetadata(e.g.,describingdatafrommultipleheterogeneoussources).4.Dataminingengine.Thisisessentialtothedataminingsystemandideallyconsistsofasetoffunctionalmodulesfortaskssuchascharacterization,associationanalysis,classification,evolutionanddeviationanalysis.5.Patternevaluationmodule.Thiscomponenttypicallyemploysinterestingnessmeasuresandinteractswiththedataminingmodulessoastofocusthesearchtowardsinterestingpatterns.Itmayaccessinterestingnessthresholdsstoredintheknowledgebase.Alternatively,thepatternevaluationmodulemaybeintegratedwiththeminingmodule,dependingontheimplementationofthedataminingmethodused.Forefficientdatamining,itishighlyrecommendedtopushtheevaluationofpatterninterestingnessasdeepaspossibleintotheminingprocesssoastoconfinethesearchtoonlytheinterestingpatterns.6.Graphicaluserinterface.Thismodulecommunicatesbetweenusersandthedataminingsystem,allowingtheusertointeractwiththesystembyspecifyingadataminingqueryortask,providinginformationtohelpfocusthesearch,andperformingexploratorydataminingbasedontheintermediatedataminingresults.Inaddition,thiscomponentallowstheusertobrowsedatabaseanddatawarehouseschemasordatastructures,evaluateminedpatterns,andvisualizethepatternsindifferentforms.Fromadatawarehouseperspective,dataminingcanbeviewedasanadvancedstageofon-1ineanalyticalprocessing(OLAP).However,datamininggoesfarbeyondthenarrowscopeofsummarization-styleanalyticalprocessingofdatawarehousesystemsbyincorporatingmoreadvancedtechniquesfordataunderstanding.Whiletheremaybemany“dataminingsystems”onthemarket,notallofthemcanperformtruedatamining.Adataanalysissystemthatdoesnothandlelargeamountsofdatacanatmostbecategorizedasamachinelearningsystem,astatisticaldataanalysistool,oranexperimentalsystemprototype.Asystemthatcanonlyperformdataorinformationretrieval,includingfindingaggregatevalues,orthatperformsdeductivequeryansweringinlargedatabasesshouldbemoreappropriatelycategorizedaseitheradatabasesystem,aninformationretrievalsystem,oradeductivedatabasesystem.Datamininginvolvesanintegrationoftechniquesfrommult1pledisciplinessuchasdatabasetechnology,statistics,machinelearning,highperformancecomputing,patternrecognition,neuralnetworks,datavisualization,informationretrieval,imageandsignalprocessing,andspatialdataanalysis.Weadoptadatabaseperspectiveinourpresentationofdatamininginthisbook.Thatis,emphasisisplacedonefficientandscalabledataminingtechniquesforlargedatabases.Byperformingdatamining,interestingknowledge,regularities,orhigh-levelinformationcanbeextractedfromdatabasesandviewedorbrowsedfromdifferentangles.Thediscoveredknowledgecanbeappliedtodecisionmaking,processcontrol,informationmanagement,queryprocessing,andsoon.Therefore,dataminingisconsideredasoneofthemostimportantfrontiersindatabasesystemsandoneofthemostpromising,newdatabaseapplicationsintheinformationindustry.AclassificationofdataminingsystemsDataminingisaninterdisciplinaryfield,theconfluenceofasetofdisciplines,includingdatabasesystems,statistics,machinelearning,visualization,andinformationscience.Moreover,dependingonthedataminingapproachused,techniquesfromotherdisciplinesmaybeapplied,suchasneuralnetworks,fuzzyandorroughsettheory,knowledgerepresentation,inductivelogicprogramming,orhighperformancecomputing.Dependingonthekindsofdatatobeminedoronthegivendataminingapplication,thedataminingsystemmayalsointegratetechniquesfromspatialdataanalysis,Informationretrieval,patternrecognition,imageanalysis,signalprocessing,computergraphics,Webtechnology,economics,orpsychology.Becauseofthediversityofdisciplinescontributingtodatamining,dataminingresearchisexpectedtogeneratealargevarietyofdataminingsystems.Therefore,itisnecessarytoprovideaclearclassificationofdataminingsystems.Suchaclassificationmayhelppotentialusersdistinguishdataminingsystemsandidentifythosethatbestmatchtheirneeds.Dataminingsystemscanbecategorizedaccordingtovariouscriteria,asfollows.1)Classificationaccordingtothekindsofdatabasesmined.Adataminingsystemcanbeclassifiedaccordingtothekindsofdatabasesmined.Databasesystemsthemselvescanbeclassifiedaccordingtodifferentcriteria(suchasdatamodels,orthetypesofdataorapplicationsinvolved),eachofwhichmayrequireitsowndataminingtechnique.Dataminingsystemscanthereforebeclassifiedaccordingly.Forinstance,ifclassifyingaccordingtodatamodels,wemayhavearelational,transactional,object-oriented,object-relational,ordatawarehouseminingsystem.Ifclassifyingaccordingtothespecialtypesofdatahandled,wemayhaveaspatial,time-series,text,ormultimediadataminingsystem,oraWorld-WideWebminingsystem.Othersystemtypesincludeheterogeneousdataminingsystems,andlegacydataminingsystems.2)Classificationaccordingtothekindsofknowledgemined.Dataminingsystemscanbecategorizedaccordingtothekindsofknowledgetheymine,i.e.,basedondataminingfunctionalities,suchascharacterization,discrimination,association,classification,clustering,trendandevolutionanalysis,deviationanalysis,similarityanalysis,etc.Acomprehensivedataminingsystemusuallyprovidesmultipleand/orintegrateddataminingfunctionalities.Moreover,dataminingsystemscanalsobedistinguishedbasedonthegranularityorlevelsofabstractionoftheknowledgemined,includinggeneralizedknowledge(atahighlevelofabstraction),primitive-levelknowledge(atarawdatalevel),orknowledgeatmultiplelevels(consideringseverallevelsofabstraction).Anadvanceddataminingsystemshouldfacilitatethediscoveryofknowledgeatmultiplelevelsofabstraction.3)Classificationaccordingtothekindsoftechniquesutilized.Dataminingsystemscanalsobecategorizedaccordingtotheunderlyingdataminingtechniquesemployed.Thesetechniquescanbedescribedaccordingtothedegreeofuserinteractioninvolved(e.g.,autonomoussystems,interactiveexploratorysystems,query-drivensystems),orthemethodsofdataanalysisemployed(e.g.,database-orientedordatawarehouse-orientedtechniques,machinelearning,statistics,visualization,patternrecognition,neuralnetworks,andsoon).Asophisticateddataminingsystemwilloftenadoptmultipledataminingtechniquesorworkoutaneffective,integratedtechniquewhichcombinesthemeritsofafewindividualapproaches.翻譯：什么是數(shù)據(jù)挖掘？簡單地說，數(shù)據(jù)挖掘是從大量的數(shù)據(jù)中提取或“挖掘”知識。該術語實際上有點兒用詞不當。注意，從礦石或砂子中挖掘黃金叫做黃金挖掘，而不是叫做礦石挖掘。這樣，數(shù)據(jù)挖掘應當更準確地命名為“從數(shù)據(jù)中挖掘知識”，不幸的是這個有點兒長?！爸R挖掘”是一個短術語，可能它不能反映出從大量數(shù)據(jù)中挖掘的意思。畢竟，挖掘是一個很生動的術語，它抓住了從大量的、未加工的材料中發(fā)現(xiàn)少量金塊這一過程的特點。這樣，這種用詞不當攜帶了“數(shù)據(jù)”和“挖掘”，就成了流行的選擇。還有一些術語，具有和數(shù)據(jù)挖掘類似但稍有不同的含義，如數(shù)據(jù)庫中的知識挖掘、知識提取、數(shù)據(jù)/模式分析、數(shù)據(jù)考古和數(shù)據(jù)捕撈。許多人把數(shù)據(jù)挖掘視為另一個常用的術語—數(shù)據(jù)庫中的知識發(fā)現(xiàn)或KDD的同義詞。而另一些人只是把數(shù)據(jù)挖掘視為數(shù)據(jù)庫中知識發(fā)現(xiàn)過程的一個基本步驟。知識發(fā)現(xiàn)的過程由以下步驟組成：1）數(shù)據(jù)清理：消除噪聲或不一致數(shù)據(jù)，2）數(shù)據(jù)集成：多種數(shù)據(jù)可以組合在一起，3）數(shù)據(jù)選擇：從數(shù)據(jù)庫中檢索與分析任務相關的數(shù)據(jù)，4）數(shù)據(jù)變換：數(shù)據(jù)變換或統(tǒng)一成適合挖掘的形式，如通過匯總或聚集操作，5）數(shù)據(jù)挖掘：基本步驟，使用智能方法提取數(shù)據(jù)模式，6）模式評估：根據(jù)某種興趣度度量，識別表示知識的真正有趣的模式，7）知識表示：使用可視化和知識表示技術，向用戶提供挖掘的知識。數(shù)據(jù)挖掘的步驟可以與用戶或知識庫進行交互。把有趣的模式提供給用戶，或作為新的知識存放在知識庫中。注意，根據(jù)這種觀點，數(shù)據(jù)挖掘只是整個過程中的一個步驟，盡管是最重要的一步，因為它發(fā)現(xiàn)隱藏的模式。我們同意數(shù)據(jù)挖掘是知識發(fā)現(xiàn)過程中的一個步驟。然而，在產業(yè)界、媒體和數(shù)據(jù)庫研究界，“數(shù)據(jù)挖掘”比那個較長的術語“數(shù)據(jù)庫中知識發(fā)現(xiàn)”更為流行。因此，在本書中，選用的術語是數(shù)據(jù)挖掘。我們采用數(shù)據(jù)挖掘的廣義觀點：數(shù)據(jù)挖掘是從存放在數(shù)據(jù)庫中或其他信息庫中的大量數(shù)據(jù)中挖掘出有趣知識的過程?；谶@種觀點，典型的數(shù)據(jù)挖掘系統(tǒng)具有以下主要成分：數(shù)據(jù)庫、數(shù)據(jù)倉庫或其他信息庫：這是一個或一組數(shù)據(jù)庫、數(shù)據(jù)倉庫、電子表格或其他類型的信息庫。可以在數(shù)據(jù)上進行數(shù)據(jù)清理和集成。數(shù)據(jù)庫、數(shù)據(jù)倉庫服務器：根據(jù)用戶的數(shù)據(jù)挖掘請求，數(shù)據(jù)庫、數(shù)據(jù)倉庫服務器負責提取相關數(shù)據(jù)。知識庫：這是領域知識，用于指導搜索，或評估結果模式的興趣度。這種知識可能包括概念分層，用于將屬性或屬性值組織成不同的抽象層。用戶確信方面的知識也可以包含在內。可以使用這種知識，根據(jù)非期望性評估模式的興趣度。領域知識的其他例子有興趣度限制或閾值和元數(shù)據(jù)（例如，描述來自多個異種數(shù)據(jù)源的數(shù)據(jù)）。數(shù)據(jù)挖掘引擎：這是數(shù)據(jù)挖掘系統(tǒng)基本的部分，由一組功能模塊組成，用于特征化、關聯(lián)、分類、聚類分析以及演變和偏差分析。模式評估模塊：通常，此成分使用興趣度度量，并與數(shù)據(jù)挖掘模塊交互，以便將搜索聚集在有趣的模式上。它可能使用興趣度閾值過濾發(fā)現(xiàn)的模式。模式評估模塊也可以與挖掘模塊集成在一起，這依賴于所用的數(shù)據(jù)挖掘方法的實現(xiàn)。對于有效的數(shù)據(jù)挖掘，建議盡可能深地將模式評估推進到挖掘過程之中，以便將搜索限制在有興趣的模式上。圖形用戶界面：本模塊在用戶和數(shù)據(jù)挖掘系統(tǒng)之間進行通信，允許用戶與系統(tǒng)進行交互，指定數(shù)據(jù)挖掘查詢或任務，提供信息、幫助搜索聚焦，根據(jù)數(shù)據(jù)挖掘的中間結果進行探索式數(shù)據(jù)挖掘。此外，此成分還允許用戶瀏覽數(shù)據(jù)庫和數(shù)據(jù)倉庫模式或數(shù)據(jù)結構，評估挖掘的模式，以不同的形式對模式進行可視化。從數(shù)據(jù)倉庫觀點，數(shù)據(jù)挖掘可以看作聯(lián)機分析處理（OLAP）的高級階段。然而，通過結合更高級的數(shù)據(jù)理解技術，數(shù)據(jù)挖掘比數(shù)據(jù)倉庫的匯總型分析處理走得更遠。盡管市場上已有許多“數(shù)據(jù)挖掘系統(tǒng)”，但是并非所有系統(tǒng)的都能進行真正的數(shù)據(jù)挖掘。不能處理大量數(shù)據(jù)的數(shù)據(jù)分析系統(tǒng)，最多是被稱作機器學習系統(tǒng)、統(tǒng)計數(shù)據(jù)分析工具或實驗系統(tǒng)原型。一個系統(tǒng)只能夠進行數(shù)據(jù)或信息檢索，包括在大型數(shù)據(jù)庫中找出聚集的值或回答演繹查詢，應當歸類為數(shù)據(jù)庫系統(tǒng)