基于情感分类的产品评论垂直有哪些信誉好的足球投注网站引擎的分析-analysis of product review vertical search engine based on emotion classification.docx

下载文档 降价啦

5
0
约5.66万字
约 72页
2018-06-05 发布于上海
举报
版权申诉
保障服务

基于情感分类的产品评论垂直有哪些信誉好的足球投注网站引擎的分析-analysis of product review vertical search engine based on emotion classification.docx

1、本文档共72页，可阅读全部内容。
2、有哪些信誉好的足球投注网站（book118）网站文档一经付费（服务费），不意味着购买了该文档的版权，仅供个人/单位学习、研究之用，不得用于商业用途，未经授权，严禁复制、发行、汇编、翻译或者网络传播等，侵权必究。
3、本站所有内容均由合作方或网友上传，本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺！文档内容仅供研究参考，付费前请自行鉴别。如您付费，意味着您自己接受本站规则且自行承担风险，本站不退款、不进行额外附加服务；查看《如何避免下载的几个坑》。如果您已付费下载过本站文档，您可以点击这里二次下载。
4、如文档侵犯商业秘密、侵犯著作权、侵犯人身权等，请点击“版权申诉”（推荐），也可以打举报电话：400-050-0827(电话支持时间：9:00-18:30)。

基于情感分类的产品评论垂直有哪些信誉好的足球投注网站引擎的分析-analysis of product review vertical search engine based on emotion classification

ABSTRACTWiththedevelopmentofInternettechnology,e-commerce,andcyber-networkproductssuchasBBS,blog,andMicroBlog,theBTOCexchangeshavebecomemoreandmorecommon.NowadaysmoreandmorecustomersputtheirproductreviewsontheInternetaftertheiruse.Likewise,thenumberofreviewsisincreasing,amongwhichmostarecolloquialandnonstructural.Itcostsmuchforbusinesspeopletoselectself-concernedinformationmanuallyfromthemassiveproductreviewstomakedecisionsaboutthesupply-demandrelations.That’sthesamecaseforpotentialcustomerstobuyproductsforitisone-sidedandlagging.ThussearchenginehasplayedasignificantroleintheInternet.ButthefamoussearchengineslikeBaiduandGooglecannotfullymeetthedemandsofresearchesinspecificproductreviews,whicharegeneralsearchenginefordifferentproductsindifferentdomains.Insuchacircumstance,thestudyanddevelopmentofaverticalsearchenginebasedontheproductreviewofclassifiedsentimentsisdefinitelynecessary.Onthebasisofresearchesandstudiesinandabroad,thisthesishasmadeafurtherstudyontherecognitionofcommenttargetincommentsofChineseproducts,evaluationphrase,collocationofcommenttargetandevaluationphrase,andontheanalysisofsentimentalorientationofevaluationphrase.Themainworkisasfollows:（1）Partofspeechsequencetemplatesareusedtoextractcandidatecommenttargetsduringtheprogressofrecognizingthecommenttargets.Thecommenttargets’integrityandstabilityandalgorithmareproposedtofiltertheuselesswordsinthecommenttargets.Thecommenttargetsandevaluationphrases’co-occurrencerulesandthefrequencyofcommenttargetsinthewholereviewareusedtosorttheconfidencelevelofcommenttargets.Andfinallycommenttargetsareextractedaccordingtotheconfidencelevel.（2）Theconjunctiondictionary,sentimentdictionary,degreewordsdictionaryandnegativewordsdictionaryareimprovedtorecognizecommentphrasesandanalysisthesentimentorientationofcommentphrases.Thereareeightfeatureswhichareabouttherelationshipbetweencommenttargetsandcommentphrases.ThesefeaturesandthealgorithmofSVMareusedtorecognizethecollocationofcommenttargetsandcommentphrases.Ultimately,thesent