资源描述：
《基于相似度加权推荐的p2p环境下的信任模型》由会员上传分享，免费在线阅读，更多相关内容在行业资料-天天文库。
1、ISSN1000-9825,CODENRUXUEWE-mail:jos@iscas.ac.cnJournalofSoftware,Vol.18,No.1,January2007,pp.157−167http://www.jos.org.cnDOI:10.1360/jos180157Tel/Fax:+86-10-62562563©2007byJournalofSoftware.Allrightsreserved.∗基于相似度加权推荐的P2P环境下的信任模型+李景涛,荆一楠,肖晓春,王雪平,张根度(复旦大学计算
2、机与信息技术系,上海200433)ATrustModelBasedonSimilarity-WeightedRecommendationforP2PEnvironments+LIJing-Tao,JINGYi-Nan,XIAOXiao-Chun,WANGXue-Ping,ZHANGGen-Du(DepartmentofComputerandInformationTechnology,FudanUniversity,Shanghai200433,China)+Correspondingauthor:Phn:+
3、86-21-51355369ext809,E-mail:lijt@fudan.edu.cn,http://www.fudan.edu.cnLiJT,JingYN,XiaoXC,WangXP,ZhangGD.Atrustmodelbasedonsimilarity-weightedrecommendationforP2Penvironments.JournalofSoftware,2007,18(1):157−167.http://www.jos.org.cn/1000-9825/18/157.htmAbst
4、ract:Indecentralizedpeer-to-peerfile-sharingnetworks,duetotheanonymousandself-organizationnatureofpeers,theyhavetomanagetheriskinvolvedwiththetransactionswithoutpriorknowledgeabouteachother’sreputation.SWRTrust,aglobaltrustmodel,isproposedtoquantifyandtoev
5、aluatethetrustworthinessofpeers,whichincludesamathematicaldescriptionandadistributedimplementation.InSWRTrust,eachpeerisassignedauniqueglobaltrustvalue,computedbyaggregatingsimilarity-weightedrecommendationsofthepeerswhohaveinteractedwithit.Previousglobalt
6、rustmodelsarebasedontheassumptionthatthepeerswithhightrustvaluewillgivethehonestrecommendation.Thispaperarguesthatthisassumptionmaynotholdinallcases.TheoreticalanalysesandexperimentalresultsshowthatSWRTrustisstillrobustundermoregeneralconditionswheremalici
7、ouspeerscooperateinanattempttodeliberatelysubvertthesystem,convergesmorequickly,anddecreasesthenumberofinauthenticfilesdownloadedmoreeffectivelythanthepreviousmodels.Keywords:peer-to-peer;trust;distributedhashtable;similarity;anonymity;convergencespeed摘要:在
8、诸如文件共享等无中心的Peer-to-Peer网络中,对等节点具有匿名性和高度自治的特点,并且由于缺乏对与之交互的节点的可信程度的知识,节点需应对交互过程中可能出现的威胁.提出了一种基于节点评分行为相似度加权推荐的peer-to-peer环境下的全局信任模型(简称SWRTrust),用于量化和评估节点的可信程度,给出了模型的数学表述和分布式实现方法.已有的全局信任模型建立在信任值高的节点其推荐也更可信这个假设基础
