Discovering Attribute-Specific Features From Online Reviews: What Is the Gap Between Automated Tools and Human Cognition?

Int. J. Softw. Sci. Comput. Intell. Pub Date : 2018-04-01 DOI:10.4018/IJSSCI.2018040101

X. Jing, Penghao Wang, Julia M. Rayz

{"title":"Discovering Attribute-Specific Features From Online Reviews: What Is the Gap Between Automated Tools and Human Cognition?","authors":"X. Jing, Penghao Wang, Julia M. Rayz","doi":"10.4018/IJSSCI.2018040101","DOIUrl":null,"url":null,"abstract":"Thisarticledescribeshowonlinereviewsplayanimportantroleindatadrivendecisionmaking. Manyeffortshavebeeninvestedindeterminingtheoverallsentimentcarriedbythereviews.However, oftentimes,theoverallratingsofthereviewsdonotrepresentopinionstowardspecificattributes ofaproduct.Anidealopinionminingtoolshouldaimatfindingboththeproductattributesand theircorrespondingopinions.Theauthorsproposeanapproachforextractingtheattributespecific featuresfromonlinereviewsusingaWord2Vecmodelcombinedwithclustering.Twoexperiments aredescribed in thispaper: thefirst focuseson testing theperformanceof theWord2Vecmodel onextractingproductaspectwords,thesecondaddresseshowwelltheextractedfeaturesobtained arerecognizablebyhumancognition.Anewmetriccalledthe“splitvalue”thatisbasedoncluster similarityanddiversityisintroducedtoexaminetheconsistencyofclusteringalgorithm.Theauthors’ experimentssuggestthatmeaningfulclusters,whichprovideinsightstotheproductattributesand sentiments,couldbeextractedfromthereviews. KeyWORDS Artificial Intelligence, Clustering, Cognition, Feature Extraction, Opinion Mining, Text Understand, Word2Vec","PeriodicalId":432255,"journal":{"name":"Int. J. Softw. Sci. Comput. Intell.","volume":"35 6 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2018-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"6","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Int. J. Softw. Sci. Comput. Intell.","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.4018/IJSSCI.2018040101","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 6

Abstract

Thisarticledescribeshowonlinereviewsplayanimportantroleindatadrivendecisionmaking. Manyeffortshavebeeninvestedindeterminingtheoverallsentimentcarriedbythereviews.However, oftentimes,theoverallratingsofthereviewsdonotrepresentopinionstowardspecificattributes ofaproduct.Anidealopinionminingtoolshouldaimatfindingboththeproductattributesand theircorrespondingopinions.Theauthorsproposeanapproachforextractingtheattributespecific featuresfromonlinereviewsusingaWord2Vecmodelcombinedwithclustering.Twoexperiments aredescribed in thispaper: thefirst focuseson testing theperformanceof theWord2Vecmodel onextractingproductaspectwords,thesecondaddresseshowwelltheextractedfeaturesobtained arerecognizablebyhumancognition.Anewmetriccalledthe“splitvalue”thatisbasedoncluster similarityanddiversityisintroducedtoexaminetheconsistencyofclusteringalgorithm.Theauthors’ experimentssuggestthatmeaningfulclusters,whichprovideinsightstotheproductattributesand sentiments,couldbeextractedfromthereviews. KeyWORDS Artificial Intelligence, Clustering, Cognition, Feature Extraction, Opinion Mining, Text Understand, Word2Vec

查看原文

微信好友朋友圈 QQ好友复制链接

本刊更多论文

从在线评论中发现属性特定的特征:自动化工具和人类认知之间的差距是什么?

Thisarticledescribeshowonlinereviewsplayanimportantroleindatadrivendecisionmaking。Manyeffortshavebeeninvestedindeterminingtheoverallsentimentcarriedbythereviews。However，通常是theoverallratingsofthereviewsdonotrepresentopinionstowardspecificattributes ofaproduct。Anidealopinionminingtoolshouldaimatfindingboththeproductattributesand theircorrespondingopinions。Theauthorsproposeanapproachforextractingtheattributespecific featuresfromonlinereviewsusingaWord2Vecmodelcombinedwithclustering。Twoexperiments aredescribed in> thispaper: > thefirst focuseson testing> theperformanceof theWord2Vecmodel onextractingproductaspectwords，thesecondaddresseshowwelltheextractedfeaturesobtained arerecognizablebyhumancognition。Anewmetriccalledthe " splitvalue " thatisbasedoncluster similarityanddiversityisintroducedtoexaminetheconsistencyofclusteringalgorithm。Theauthors ' ' experimentssuggestthatmeaningfulclusters，whichprovideinsightstotheproductattributesand情绪，couldbeextractedfromthereviews。关键词:人工智能，聚类，认知，特征提取，意见挖掘，文本理解，Word2Vec

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文去求助

来源期刊

Int. J. Softw. Sci. Comput. Intell.

自引率

0.00%

发文量