用戶興趣模型在RSS網(wǎng)絡(luò)信息服務(wù)中的研究.pdf

1、南京航空航天大學(xué)碩士學(xué)位論文用戶興趣模型在RSS網(wǎng)絡(luò)信息服務(wù)中的研究姓名：張艷申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：計(jì)算機(jī)科學(xué)與技術(shù)指導(dǎo)教師：章勇201103用戶興趣模型在RSS網(wǎng)絡(luò)信息服務(wù)中的研究IIABSTRACTPopularityoftheInterhasbroughttheproblemofinfmationoverload.AlthoughtheRSSinfmationservicesallowsinfmationpushservices

2、cansolvetheproblemofinfmationoverloadatacertainextentbutitlackspersonalizedservices.Asthebasisceofpersonalizedinfmationserviceusermodelingtechnologyisoneofthefocusesofcurrentresearch..Indertobetterrealizethepersonalizedi

3、nfmationservicesthispaperhasanindepthstudyofthetechnologyrelatedtotheuserinterestmodelappliesittoRSSinfmationservices.Thewkcontributionofthispaperisasfollows:Firstthispaperimprovestheexpresswayofuserinterestmodelbasedont

4、hevectspacemodeladdinganidentifyoflongtermshtterminteresttheupdatedtimealsoimprovesthetraditionalwayofTFIDFfmulatocalculatetheweightofkeywdsaddingpageinterestgradeparameters.Theimprovedmethodismeeffectivetoexpresstheuser

5、’sinterest.SecondthispaperproposesanewalgithmcombininggradualfgettingwithRocchiomethodtoupdatethelongterminterestmodel.Theshtterminterestcanbeupdatedwiththeslidingtimewindowitcanchangeintolongterminterestwhensatisfyingce

6、rtainconditions.Theproposedmodelupdatingalgithmcanbetimelyaccuratelytracktheuser’sinterestdrift.Thirdthispaperproposesanoptimizationalgithmofuserinterestmodelbasedoninfmationrecommendationtechnology.Thealgithmcanfurthero

7、ptimizethemixeduserinterestmodelcanadaptivelyadjusttheproptionofthelongtermshtterminterestintheinfmationrecommendation.Italsocanbetterrecommendresourcesfusers.ThenthispaperdesignsimplementspersonalizedRSSnewsrecommendati

8、onsystemaccdingtotheupdateoptimizationalgithmofthemodelproposedinthispaper.ThesystemcancombineRSStechnologywithpersonalizedinfmationservices.Theexperimentalresultsbasedonthesystemshowthattheupdateoptimizationalgithmisfea