I am deeply indebted to Dr. Michael Berry, my major advisor, for his kind guidance and support. I also thank Dr. Susan Dumais, director of the Information Sciences Research Group at Bellcore, for her technical advice. In addition, she graciously allowed us
Chapter2
Vector-SpaceModelsforInformationRetrieval
Researchininformationretrievalhasfollowedseveralparallel,yetsimilar,http://www.77cn.com.cntentSemanticIndexing(LSI),becauseofthewayitrepresentstermsanddocumentsinaterm-documentspace,isconsideredavector-spaceinformationretrievalmodel.Inthefollowingsections,generalvector-spacemodelsareintroducedandvariousimprovementstothemodelsarediscussed.Inthe nalsection,LSIitselfisconsidered.
2.1IntroductiontoVector-SpaceModels
Thevector-spacemodelsforinformationretrievalarejustonesubclassofretrievaltechniquesthathavebeenstudiedinrecentyears.Thetaxonomyprovidedin[BC87]labelstheclassoftechniquesthatresemblevector-spacemodels“formal,feature-based,individual,partialmatch”retrievaltechniquessincetheytypicallyrelyonanunderlying,formalmathematicalmodelforretrieval,modelthedocumentsassetsoftermsthatcanbeindividuallyweightedandmanipulated,performqueriesby
6