欢迎来到天天文库
浏览记录
ID:34188650
大小:1.69 MB
页数:56页
时间:2019-03-03
《基于个人微博时序事件的-研究》由会员上传分享,免费在线阅读,更多相关内容在教育资源-天天文库。
1、内蒙古科技大学硕士学位论文category.Atthesametime,computedTF-IDFvaluestoextractmicro-blogfeaturewordsaftermakingthefeaturewordsconformeduniformdistributioninthetext;ThenThen,OnthebasisoftheextractedfeaturewordsandtermsofsimilarstructureonthetemplatecombineJaccardsimilarityandcosin
2、esimilaritytocalculateComprehensivesimilaritycalculationofpersonalmicro-blog,Thealgorithmovercomestheinadequateoftraditionalapproachbasedonlywordco-occurrence,fromsimilarwordsfeaturesandcharacteristicsoftheindividualvaluesandotheraspectstocalculatethesimilarityofthet
3、womicro-blogdeeplyandcomprehensively;Finally,utilizedK-Meansalgorithmtoclassifytimingeventsofindividualsmicro-blog,makedthesametopicmicro-blogclassifiedintothesamecollection.Theexperimentalresultsshowthattheproposedcomprehensivesimilarityalgorithmsalgorithmhashigherp
4、recisionthanthetraditionalsimilarity.Tosomeextent,itimprovedaccuracyoftimingeventsclassifiedofpersonalmicro-blog.Keywords:Featurewords;Similarwordtemplate;Similarity;EventclassificationIII内蒙古科技大学硕士学位论文目录摘要..............................................................
5、......................................................................IAbstract.....................................................................................................................................II1绪论..................................................
6、.....................................................................................11.1研究背景.......................................................................................................................11.1.1微博...............................................
7、.........................................................................21.1.2微博时序事件挖掘...............................................................................................31.2国内外研究现状..........................................................................
8、.................................31.3论文主要研究内容......................................................................................
此文档下载收益归作者所有