资源描述:
《基于微博的突发事件检测和信息传播建模》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库。
1、AbstractAbstractMicroblogsarebelovedbyuserssinceitsadventbecauseofitstimelinessandconvenience,andbecomeusers’favoriteamongthenetworkapplications.Microblogbecomesapowerfultoolforobservingthesocietywithitsmassiveamountofusersandrelativelyfreespeech.Thispaperfocusesonthedetectionofemergenteventonmicrob
2、logsandthemodelingofinformationdiffusionthroughmicroblognetworks,basedonthemicroblogplatform.First,weobtainthedatasetformicroblogstudy.Thispapermakeaninformationextractionprocesson"renminmicroblog",usingthecrawlerdesignedaccordingtothepageinformationstructureofthemicroblogplatform,thenextractandstor
3、etheoverallinformationandhotinformationoftheplatform.Theoverallone-timeinformationextractionincludesmicrobloginformation,userinformationanduserrelationshipinformation.Hotreal-timeinformationextractionincludeshotmicrobloginformation,hotuserinformationandhotkeyinformation.Then,wedetecttheemergentevent
4、sinmicroblogs.Thebriefnessofmicroblogrendersthatinitsdescriptionoftheevent,thekeywordsoftheeventsappearinhighfrequency,andpresentedwithconsistentconcerndegreeoftheevents.Thus,theemergenteventdetectioninthispaperfollowsthedetectionofkeywordsoftheevents.First,methodsoffeaturevalueselectionanddataorgan
5、izationformicroblogcorpusarepresented.Then,thefeaturetrajectoryofeachexpectedwordisconstructed,itstimedomainandfrequencydomainfeaturecanjudgetheburstinessoftheword.Inthisprocess,thediscoveryofnewwordsisexecuted,becausethekeywordsofthebursteventsincludeinformationofthepeopleandtheplaceinformation,and
6、thesewordsareusuallyunknownwordsbutofgreatsignificancetotheexpressionoftheevents.Finally,thewordsareclusteredaccordingtotheconcurrenceinmicroblogs,realizingthecorrespondenceofburstkeywordsandevents,andthedetaileddescriptionofthebursteventisprovidedusingtheobtainedinformationfrommicroblogs.Finally,we
7、constructaninformationdiffusionmodelforthenetworkofmicroblogusers.Thispaperconstructsafive-tuplemodelofinformationdiffusioninmicroblog,combiningtherouteandthecharacteristicsofinformationdiffusionin-II