Springer.Data.Mining.eBook.Springer.Survey.of.Text.Mining.Clustering.Classification.and.Retrieval.[Scan.OCR].2004.(By.Laxxuss)

《Springer.Data.Mining.eBook.Springer.Survey.of.Text.Mining.Clustering.Classification.and.Retrieval.[Scan.OCR].2004.(By.Laxxuss)》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、SpringerNewYorkBerlinHeidelbergHongKongLondonMilanParisTokyoMichaelW.BerryEditorSurveyofTextMiningClustering,Classification,andRetrievalScannedbyVelocityWith57IllustrationsSpringerMichaelW.BerryDepartmentofComputerScienceUniversityofTennessee203ClaxtonComplex

2、Knoxville,TN37996-3450,USAberry@cs.utk.eduCoverillustration:VisualizationofthreemajorclustersintheL.A.Timesnewsdatabasewhendocumentvectorsareprojectedintothe3-DsubspacespannedbythethreemostrelevantaxesdeterminedusingCOVrescale.Thisfigureappearsonp.118ofthetex

3、t.LibraryofCongressCataloging-in-PublicationDataSurveyoftextmining:clustering,classification,andretrieval/editor,MichaelW.Berry.p.cm.Includesbibliographicalreferencesandindex.ISBN0-387-95563-1(alk.Paper)1.Datamining—Congresses.2.Clusteranalysis—Congresses.3.D

4、iscriminantanalysis—Congresses.I.Berry,MichaelW.QA76.9.D343S692003006.3—dc212003042434ISBN0-387-95563-1Printedonacid-freepaper.©2004Springer-VerlagNewYork,Inc.Allrightsreserved.Thisworkmaynotbetranslatedorcopiedinwholeorinpartwithoutthewrittenpermissionofthep

5、ublisher(Springer-VerlagNewYork,Inc.,175FifthAvenue,NewYork,NY10010,USA),exceptforbriefexcerptsinconnectionwithreviewsorscholarlyanalysis.Useinconnectionwithanyformofinformationstorageandretrieval,electronicadaptation,computersoftware,orbysimilarordissimilarm

6、ethodologynowknownorhereafterdevelopedisforbidden.Theuseinthispublicationoftradenames,trademarks,servicemarks,andsimilarterms,eveniftheyarenotidentifiedassuch,isnottobetakenasanexpressionofopinionastowhetherornottheyaresubjecttoproprietaryrights.987654321SPIN

7、10890871www.springer-ny.comSpringer-VerlagNewYorkBerlinHeidelbergAmemberofBertelsmannSpringerScience+BusinessMediaGmbHPrefacexiContributorsxiiiIClusteringandClassification11Cluster-PreservingDimensionReductionMethodsforEfficient(ClassificationofTextData3PegHo

8、wlandandHaesunPark1.1Introduction31.2DimensionReductionintheVectorSpaceMode)41.3AMethodBasedonanOrthogonalBasisofCentroids51.3.1RelationshiptoaMethodfromFactorAnalysis....71.4Discriminant