資源描述：
《基于標簽協(xié)同過濾算法在微博推薦中的分析》由會員上傳分享，免費在線閱讀，更多相關內(nèi)容在學術論文-天天文庫。
1、內(nèi)蒙古科技大學碩士學位論文AbstractWiththevigorousdevelopmentofWeb2.0technology,theglobalinternethasenteredthenetworkinteractiveage.ThebirthanddevelopmentofmicroblognetworkhasboughtenormousinfluencetothetransmittingstyleoftheInternetandusers'everydaylife.Userscanlistentoanytopicthattheylik
2、e,beclosedtodailylifeandprofessionalfieldbyusingmicroblogplatform.Withtheriseofthemicroblogusers,themicrobloginformationalsomultipliedgrowth.Findingtheuserswhohavecommoninterestswithyouinmanymicroblogusersisjustwanttogetinformation.Therefore,recommendingfriendswithcommontaste
3、stomicroblogusershasbecomethefocusofthispaper.Thispaperstudiedtheexistingsocialnetworkfriendrecommendationalgorithm,throughsummarizinglearningrecommendationalgorithmtheoryknowledge,andcombinedwiththecharacteristicsofmicroblogusersfriends,agoodfriendrecommendationalgorithmbase
4、donassociationrulesandtagpersonalizedwasproposed,whichrecommendedthemostsimilarusertotargetedcustomerforitsgoodfriends.Firstly,relatedconceptsoffriendsrecommendinmicroblogweredefined,andthefriendsrecommendsystemimplementationprocesswasintroduced;secondly,calculationmethodsabo
5、utcommonfriendsrelationshipbetweenuserswerepresented,whichwasbasedonassociaterulealgorithm,andthencalculationmethodsaboutthesimilaritybetweentheuserwerepresented,whichwasbasedontagsimilarityalgorithm;finally,co-mbiningwithcommonfriendsrelationshipandtagsimilarity,thecalculati
6、onmethodbasedonpersonalizedfriendsrecommendedwasdeduced.Intheendoftheexperiment,firstthemicroblogpersonalizedfriendsrecommendsystemwasdesigned,whichprovidedexperimentalplatformsupportforperformancetestofthealgorithm.Thenweightvaluetestandalgorithmperformancetestwereca-rriedou
7、tbyusinggoodfriendsrecommendsystemrespectively.Thereinto,theweightvalueexperimentalresultsshowthatpersonalizedrecommendationalgorithmattainedtotheoptimumwhentheweightvaluewas0.6.Atthelast,comparedwiththeexistingthreefriendsalgorithm,theoptimizationsuitabilityofpersonalizedfri
8、endsrecom-mendationalgorithmwasderived,whichindicatedcombinedpersona
