improving collaborativefiltering with trust

improving collaborativefiltering with trust

ID:7294143

大小:186.55 KB

页数:5页

时间:2018-02-10

improving collaborativefiltering with trust_第1页
improving collaborativefiltering with trust_第2页
improving collaborativefiltering with trust_第3页
improving collaborativefiltering with trust_第4页
improving collaborativefiltering with trust_第5页
资源描述:

《improving collaborativefiltering with trust》由会员上传分享,免费在线阅读,更多相关内容在工程资料-天天文库

1、ImprovingCollaborativeFilteringwithTrust-basedMetricsJianshuWengChunyanMiaoAngelaGohSchoolofComputerEngineering,NanyangTechnologicalUniversity,Singapore{WENG0004,ASCYMiao,ASESGoh}@ntu.edu.sgABSTRACTitems’ratings,leadingtolowcoverage.•Similarity-basedCFisusuallysusceptibletoattacks.Des

2、piteitssuccess,similarity-basedcollaborativefilteringForexample,asthenumberofsimilarusersisusuallysuffersfromsomesignificantlimitations,suchasscalability1andsparsity.Thispaperintroducestrusttothedomainofsmall,malicioususers’noisyratingscaneasilyout-collaborativefilteringtoovercometheselim

3、itations.Com-weighotherusers’ratings.Asaresult,presenceofnoisyratingswillbiasthepredictions.paredwiththesimilarity-basedCF,introductionoftrustThispaperpresentsoureffortstoovercometheselimita-doesimprovetheperformanceofCFintermsofcoverage,tions.predictionaccuracy,androbustnessintheprese

4、nceofat-Itisproposedthateachusermaintainsratingsindivid-tacks.Experimentalresultsbasedonarealdatasetareil-ually.Eachuseristreatedequallyasapeerandconnectslustratedasevidencestosupportourclaim.witheachothertoformacollaborativeinformationfilteringCategoriesandSubjectDescriptors:H.3.3[Inf

5、orma-community,whichisinfactaP2Poverlaynetwork.ThetionSearchandRetrieval]:InformationfilteringP2P-basedarchitectureisthenexploitedtofacilitateusers’GeneralTerms:Algorithm,Performancerecommendationscollection.TheP2P-basedarchitectureexemptstheneedforacentralizedserver,leadingtohigherKey

6、words:CollaborativeFiltering,Trustscalability.1.INTRODUCTIONFurthermore,atrustmetricsisdesignedwhichhelpsaCollaborativefiltering(CF),whichpredictsnewitems’usertoquantifythedegreeoftrustitshouldplaceonaratingsforactiveusersbasedonotherusers’ratings,isoneparticularuser.Thetrustmetricsisc

7、omputableonmostofthemostsuccessfultechniquesontheweb.Traditionally,users,evenonpairsofuserswhohaveonlyco-ratedoneCFissimilarity-based,i.e.thepredictionsaremadebasedcommonitem.Auserisalsoabletoestablishtrustviatrustonsimilarusers’ratings[8].Theuser-ratingdataisusuallypropagationonusers

8、withw

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。