原创力文档- 1、原创力文档（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。。
- 2、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。
- 3、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)。
- 4、该文档为VIP文档，如果想要下载，成为VIP会员后，下载免费。
- 5、成为VIP后，下载本文档将扣除1次下载权益。下载后，不支持退款、换文档。如有疑问请联系我们。
- 6、成为VIP后，您将拥有八大权益，权益包括：VIP文档下载权益、阅读免打扰、文档格式转换、高级专利检索、专属身份标志、高级客服、多端互通、版权登记。
- 7、VIP文档为合作方或网友上传，每下载1次， 网站将根据用户上传文档的质量评分、类型等，对文档贡献者给予高额补贴、流量扶持。如果你也想贡献VIP文档。上传文档
查看更多
Sentiment Analysis with Data-driven ConstraintsA/Prof. Minlie Huang （黄民烈）Dept. of Computer Science TechnologyTsinghua University?
Our work on opinion miningSpam/Spammer DetectionOpinion ExtractionOpinion classification+-IJCAI 2011CIKM 2015ACL 2014EMNLP 2014COLING 2010Opinion SummarizationJCST 2011COLING 2010ACL 2015IJCAI 2015CIKM 2014CIKM 2013ACL 2012AAAI 2010Opinion RetrievalACL 2009Sentiment RatingKAIS 2014
Unsupervised or Semi-supervised Learning for Sentiment Analysis?Lack of direct supervision?Maximizing likelihood may not give expected output ...We have：train：unlabeled datalittle or no labeled datato maximize likelihood of observed data
imple statistical assumption+large-scale unlabeled dataTraining Processindirect supervision data-driven constraints
Data-driven Constraints in Sentiment AnalysisSentimentDistributionConsistencySyntacticDistributionConsistencyWord ClassDistributionConsistencysimple statistical assumption+ large-scale unlabeled
Data-driven Constraints in Sentiment AnalysisWord Class Distribution ConsistencySentiment\Text ClassificationSentiment Distribution ConsistencyClustering Aspect\Opinion WordsSyntactic Distribution Consistencysimple statistical assumptisoentiment analysis tasks
Background
Clustering Aspect\Opinion Words?Goal：clustering aspect phrases that refer to the same product property?aspect “battery”: {“battery”, “battery life”, “power”…}
Sentiment\Text Classification?Goal：classify an document as expressing a positive or negative opinion or sentiment.
Sentiment Distribution ConsistencyLi Zhao, Minlie Huang, Haiqiang Chen, Junjun Cheng, Xiaoyan Zhu.Clustering Aspect-related Phrases by Leveraging Sentiment Distribution Consistency. EMNLP 2014, October 25–29, 2014 — Doha, Qatar.
Semi-structured Review Data
Sentiment Distribution ConsistencyThe sentiment distribution of aspect “battery” and its related-phrases on A with a large amount of reviews.
Sentiment Distribution ConsistencySentiment Distribution Consistency Assumption: Different p
文档评论（0）