国外计算机科学类论文,推荐系统
332W.-P.Lee,T.-H.Yang/ExpertSystemswithApplications25(2003)331–341
environmentsuchastelevision,anidealsolutiontotheuseristohaveapersonalidenti er(liketheSIMcardinthemobilephone)thatcanrecordhismostrecentpreferencesandworkasa lterforchannelselection.Therefore,wheneverauserequipshisidenti erintoaspecialdevice(e.g.remotecontrol)toturnonthetelevision,theprogrammesavailableatthattimearesortedautomaticallyandpresentedtohimbasedonhispreferences.Inotherwords,the rstprogrammepresentedtoauserispresumedtobetheonehelikesthemostamongtheprogrammescurrentlybroadcasted;thenextoneisthesecondpreferred,andsoon.Inthisway,theaudiencecanimmediatelyaccesstheprogrammesheisinterestedinandthensavemuchofthetimespentonreadingtheinformationaboutthecontentsorbrowsingthechannelstodecidewhattowatch.
Inthispaper,wedescribehowweadoptanagent-basedmethodologytodevelopaprototypesystemforpersonalizedTVprogrammerecommendations.Twotypesofprogrammesareconsidered:oneisthe lm,andtheother,thenews.Todealwithdifferentprogrammes,oursystemincludes veagents:a lmagenttoimplicitlyobserveauser’ssur ngbehaviorandthenrecordwhathewatches,ortogatherauser’spreferencesinanexplicitway;anewsagenttocollectInternetnewsforuser’sidenti cation;alearningagenttoanalyzeandmodelauser’sinterests;acollaborationagenttolookforotheruserswithsimilartastesforsharingtheinformationcollected;andanadaptationagenttoevaluatetheperformanceofthesystemandtoadapttotheuser’snewlychangedinterests.Theexperimentalresultsshowthatourlearningagentcanlearnauser’sinterestsinrealtimeandgivesensiblerecommendations.Inaddition,thek-nearestneighbormethodisimplementedasabaselineforperformancecomparison.Furthermore,wealsoshowthatfordifferentsituationsinwhichauserchoosesaprogrammewithlowerpriorityratherthanthetoponesuggestedbythesystem,thelearningagentcanberesumedaccordinglyanditcontinuouslylearnsthemostup-to-datepreferencesfortheuser.
2.Background
Ingeneral,therearetwokindsofapproachesindevelopingarecommendersystem:personalized(content-based)andsocialized(collaborative)ones.Inaperson-alizedsystem,auser’spersonalinformationis rstcollected,thenthesystemreasonstheuser’spreferencesbyanalyzingandmodelingthepersonalinformationavailable.Theensuinguserpro leisrecordedbythesystemforfurtherrecommendations.Tocollectinformation,thesystemcanaskausertomanuallyindicatehisinterestsbygivingsomeexampleswithpersonalratings.Alternativelythesystemcanimplicitly
observeandrecordtheuser’soperatingbehavior,andthenanalyzesthecontenttheuserhasreadinordertobuildamodeltodistinguishbetweenwhathelikesandwhathedoesnot.
Themostimportantissueinthepersonalizationworkistoconstructacomputationalmodelforeachindividualusertopredicthispreferencesforincominginformation.Withtheusermodelworkingasa lter,onlytheinformationconsideredasuser-interestedcanreachtheenduser.Therefore,theworkofrecommendationcanberegardedasclassi cation:usingtheinformationalreadyknowntobuildamodeltopredicttheeventsunknown.Themodelmaybeasetofrules,adecisiontree,orasetofnumericalnumbersrepresentingthecorrespondingweightsforsomespeci cfeatures.Somemachinelearningapproacheshavebeenappliedtoconstructusers’pro les.InBalabanovicandShoham(1997),astatistic-basedapproacht df(termfrequencyandinversedocumentfrequency(Salton,1989))wasusedtobuildauser’spro letorecommendWebpages.InJoachims,Freitag,andMitchell(1997),areinforcementlearningmethodwasemployedforWebpagerecommendations,andinMooneyandRoy(2000),forbookrecommen-dations.Inthiswork,weuseadecisiontree-basedapproachtodevelopalearningagenttomodelauser’sinterestsforTVprogrammeselections,andtoexploredifferentprogrammefeaturesforbetterrecommendations.Insteadofbuildingapersonalmodeltopredicthowauserlikesanitem,acollaborativesystemrecommendsitemstoaspeci cuseraccordingtotheevaluationsfromotheruserswithsimilartastes.Inotherwords,thesystemdoesnotanalyzewhatauserlikesbuttakingtheopinionsofothers.Insuchasystem,thesimilarityoftheusersismeasuredbycertaincorrelationcriterion,andthenak-nearestneighbormethodisusedto ndsomeuserswithsimilarinterestsforaspeci cuser.Thepredictionofanunknownitemforauseristhusbasedonthecombinationoftheratingsofhisnearestneighbors.Goodetal.(1999)andShardanandandMaes(1995)presentsystemsofthiskind.
Somehybridmethodsareproposedtoovercometheproblemswiththeabovetwokindsofrecommendersystems,forexample,thecombinedtechniquesforpersonalizationandcollaborationusedinBalabanovicandShoham(1997),Basu,Hirsh,andCohen(1998),andSmythandCotter(2000).InSmythandCotter(2000),apersonalpro leiscompiledfromtheTVguideforprogrammerecommendations.Ourworkpresentedherediffersfromtheirsnotonlyinthelearningapproachused,butalsointhesituationsoftwodifferenttypesofprogrammesbroadcasted(i.e. lmsandnews).Mostimportantly,ourworkincludesaspeciallydesignedmechanismforsystemadaptation,withwhichauser’spro lecanbere-learnttocorrectthepredictionerrorsandtomeettheuser’smostrecentpreferences.
A MULTI-AGENT BASED SIMULATION FRAMEWORK FOR THE STUDY OF HUMAN AND ...The database module is to maintain all the information about the physical ...
A Framework for Multi-Ag... 暂无评价 10页 免费 A Multi-Agent ...Information Systems in the School of ¨ Business and Economics of the Freie...
Multi-agent framework for third party logistics in E-commerce Multi-agent ...The e-commerce system of the 3PL vendor could provide the information ...
framework for analysis and design of multi-agent supply chain systems_六...Lin and Shaw (1998) propose a multi-agent information sys- R. Govindu, ...
An Abductive Framework f... 暂无评价 17页 免费 A Multi-Agent Location ... Failure semantics for the exchange of information in multi-agent systems ...
2009 Third International Symposium on Intelligent Information Technology Application Consensus of Multi-agent Systems with Sampled Information and Input Delay ...
Simulation Framework for the Validation of Agent-based and Multi-Agent Sys...Information and Software Technology, 46(13), pp. 907-921, Oct. 2004. [...
USING MULTI-AGENT SYSTEM... 暂无评价 8页 免费 A Framework for Multi-Ag...The thesis also investigates how real-time information may be used to ...
SHADE [4] is primarily concerned with the information sharing aspect of the...Tenenbaum, An intelligent agent framework for enterprise integration, IEEE ...
that local data pre-processing decreases the information quality within the ...tributed Simulation Framework for Ambient Networks. Recent Advances in ...
热门文档