Toward Large-Scale Information Retrieval Using Latent Semant(17)

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

Vector-spacemodelsweredevelopedtoeliminatemanyoftheproblemsassociatedwithexact,lexicalmatchingtechniques.Inparticular,sincewordsoftenhavemultiplemeanings(polysemy),itisdif cultforalexicalmatchingtechniquetodifferentiatebetweentwodocumentsthatshareagivenword,butuseitdifferently,withoutun-derstandingthecontextinwhichthewordwasused.Also,sincetherearemanywaystodescribeagivenconcept(synonomy),relateddocumentsmaynotusethesameterminologytodescribetheirsharedconcepts.Aqueryusingtheterminologyofonedocumentwillnotretrievetheotherrelateddocuments.Intheworstcase,aqueryusingterminologydifferentthanthatusedbyrelateddocumentsinthecollectionmaynotretrieveanydocumentsusinglexicalmatching,eventhoughthecollectioncontainsrelateddocuments[BDO95].

Vector-spacemodels,byplacingterms,documents,andqueriesinaterm-documentspaceandcomputingsimilaritiesbetweenthequeriesandthetermsordocuments,al-lowtheresultsofaquerytoberankedaccordingtothesimilaritymeasureused.Unlikelexicalmatchingtechniquesthatprovidenorankingoraverycruderankingscheme(forexample,rankingonedocumentbeforeanotherdocumentbecauseitcon-tainsmoreoccurrencesofthesearchterms),thevector-spacemodels,bybasingtheirrankingsontheEuclideandistanceortheanglemeasurebetweenthequeryandtermsordocumentsinthespace,areabletoautomaticallyguidetheusertodocumentsthatmightbemoreconceptuallysimilarandofgreaterusethanotherdocuments.Also,byrepresentingtermsanddocumentsinthesamespace,vector-spacemodelsoftenprovideanelegantmethodofimplementingrelevancefeedback[SB90].Relevancefeedback,byallowingdocumentsaswellastermstoformthequery,andusingthetermsinthosedocumentstosupplementthequery,increasesthelengthandprecisionofthequery,helpingtheusertomoreaccuratelyspecifywhatheorshedesiresfromthesearch.

Informationretrievalmodelstypicallyexpresstheretrievalperformanceofthesystemintermsoftwoquantities:precisionandrecall.Precisionistheratioofthenumberofrelevantdocumentsretrievedbythesystemtothetotalnumberof

8

百度搜索“77cn”或“免费范文网”即可找到本站免费阅读全部范文。收藏本站方便下次阅读，免费范文网，提供经典小说综合文库Toward Large-Scale Information Retrieval Using Latent Semant(17)在线全文阅读。