文档介绍:
武汉理工大学硕士学位论文AbstractIntemetbringsconveniencetopeople.Howcvcr,isgiant,andisgrowingrapidly.So,sometimesitisnoteasytofindtheinformationneededfromthedata嬲sea.Asearchengineappears,whichbringsconveniencetopeople.Thetraditionalsearchengineismature,mercially.Howcvel',thetraditionalsearchengineisgenerallybased011matchingkeywords,andlittleusers’personalizedinformationisconsidered,whichislackofpersonalizedfeature.Inordertosolvetheproblem,apersonalizedsearchengineispresentedtomakeUSCofthepersonalizedinformation,whichhaspersonalizedfeatureandcanmeetusers’personalizedinformationneed.Inthisthesis,thepersonalizedsearelaengineisbased011thetraditionalsearchengine,andpersonalizedfeatureisaddedtomeetusers’personalizedinformationneed.ThepersonalizedsearchengineUSeSavectorspacemodel嬲auserinterestmodeltospecifyusers’interestandbehavior.Inthepersonalizedsearchengine,thouserinterestmodelisbuilt,updatedandusedtooptimizequeriesandsortresults,whichcallrealizethepersonalization.Theuserinterestmodelisthekeytorealizethepersonalizationforthepersonalizedsearchengine,andisthekeyoftheresearchinthisthesisaswell.TheUSerinterestmodelispresentedasavectorwhichconsistsoftermsandtheirweights.ThetcrllaSandtheirweightscanspecifyusers’interestandbehavior,which啪begotinsuchways鹤someinterestkeywordscanbeprovided鹪termswhenaltlscrregisters,termscanbeexlxaetedfromthequeriesinputbyauser,andtermsc龃beextractedfromthewebdocumentsbrowsedbyaUSerinthewayoftheusereliek.Getusers’interestandbehavior,andbuildtheuserinterestmodel.InordertospeCifyUSeI'5’interestandbehaviorbetter,theuscrinterestmodelneedbeupdated.Theuserinterestmodelcarlbeusedtooptimizequeriesandsortresults.Tooptimizequeries,forauser,ordingtotheirweights,sometermswithgreaterweights锄.elistedfortheusertoselect,andtheselectedtermsareusedforretrievingwiththequeriesinputbytheUSertogether.Tosortresults,forauser,calculatethesimilarityofeverywebdocumentIlsearchenginePSEmakesuseofUSerS’personalizedinformation,whichcanmeetusers’personalizedinformationneedbybuilding,updatin
内容来自淘豆网www.taodocs.com转载请标明出处.