manipulation-resistant reputations system

manipulation-resistant reputations system

ID:7290214

大?。?60.37 KB

頁數(shù):11頁

時(shí)間:2018-02-10

manipulation-resistant reputations system_第1頁
manipulation-resistant reputations system_第2頁
manipulation-resistant reputations system_第3頁
manipulation-resistant reputations system_第4頁
manipulation-resistant reputations system_第5頁
資源描述:

《manipulation-resistant reputations system》由會(huì)員上傳分享,免費(fèi)在線閱讀,更多相關(guān)內(nèi)容在工程資料-天天文庫。

1、Manipulation-ResistantReputationsUsingHittingTimeJohnHopcroftDanielSheldonCornellUniversityCornellUniversityjeh@cs.cornell.edudsheldon@cs.cornell.eduAbstractPopularreputationsystemsforlinkednetworkscanbemanipulatedbyspammerswhostrategicallyplacelinks.Thereputationofnodevisinterpretedasthewo

2、rld’sopinionofv’simportance.InPageRank[4],v’sownopinioncanbeseentohaveconsiderablein?uenceonherreputation,wherevexpressesahighopinionofherselfbyparticipatinginshortdirectedcycles.Incontrast,weshowthatexpectedhittingtime—thetimetoreachvinarandomwalk—measuresessentiallythesamequantityasPageRan

3、k,butexcludesv’sopinion.Wemakethesenotionsprecise,andshowthatareputationsystembasedonhittingtimeresiststamperingbyindividualsorgroupswhostrategicallyplaceoutlinks.Wealsopresentanalgorithmtoef?cientlycomputehittingtimeforallnodesinamassivegraph;conventionalalgorithmsdonotscaleadequately.1Intr

4、oductionReputationandrankingsystemsareanessentialpartofwebsearchande-commerce.Thegeneralideaisthatthereputationofoneparticipantisdeterminedbytheendorsementsofothers;forexam-ple,onewebpageendorsesanotherbylinkingtoit.However,notallparticipantsarehonorable—e.g.,spammerswilldotheirbesttomanipul

5、ateasearchengine’srankings.Anaturalrequirementforareputationsystemisthatindividualsshouldnotbeabletoimprovetheirownreputationus-ingsimpleself-endorsementstrategies,likeparticipatinginshortcyclestoboostPageRank.SincePageRankenjoysmanyniceproperties,itisinstructivetoseewherethingsgowrong.LetG=

6、(V;E)beadirectedgraph(e.g,theweb).PageRankassignsascore(v)toeachnodev,whereisde?nedtobethestationarydistributionofarandomwalkonG,givingthepleasinginterpretationthatthescoreofpagevisthefractionoftimeawebsurferspendsthereifsherandomlyfollowslinksforever.Fortechnicalreasons,therandomwalkismod

7、i?edtorestartineachstepwithprobability,jumpingtoapagechosenuniformlyatrandom.Thisensuresthatexistsandisef?cienttocompute.Thenawell-knownfactaboutMarkovchains[1]saysthat1=(v)isequaltotheexpectedreturntimeofv,thenumberofstepsittakesarandomwalkstart

當(dāng)前文檔最多預(yù)覽五頁,下載文檔查看全文

此文檔下載收益歸作者所有

當(dāng)前文檔最多預(yù)覽五頁,下載文檔查看全文
溫馨提示:
1. 部分包含數(shù)學(xué)公式或PPT動(dòng)畫的文件,查看預(yù)覽時(shí)可能會(huì)顯示錯(cuò)亂或異常,文件下載后無此問題,請(qǐng)放心下載。
2. 本文檔由用戶上傳,版權(quán)歸屬用戶,天天文庫負(fù)責(zé)整理代發(fā)布。如果您對(duì)本文檔版權(quán)有爭(zhēng)議請(qǐng)及時(shí)聯(lián)系客服。
3. 下載前請(qǐng)仔細(xì)閱讀文檔內(nèi)容,確認(rèn)文檔內(nèi)容符合您的需求后進(jìn)行下載,若出現(xiàn)內(nèi)容與標(biāo)題不符可向本站投訴處理。
4. 下載文檔時(shí)可能由于網(wǎng)絡(luò)波動(dòng)等原因無法下載或下載錯(cuò)誤,付費(fèi)完成后未能成功下載的用戶請(qǐng)聯(lián)系客服處理。