首页 | 本学科首页   官方微博 | 高级检索  
     

文本倾向性分析综述
引用本文:厉小军,戴霖,施寒潇,黄琦. 文本倾向性分析综述[J]. 浙江大学学报(工学版), 2011, 45(7): 1167-1174. DOI: 10.3785/j.issn.1008-973X.2011.07.005
作者姓名:厉小军  戴霖  施寒潇  黄琦
作者单位:1.浙江工商大学 计算机与信息工程学院,浙江 杭州 310018; 2.浙江大学 计算机科学与技术学院,浙江 杭州 310027
基金项目:浙江省重大科技专项资助项目(2008C13082);浙江省自然科学基金资助项目(Y1090688);中央高校基本科研业务费专项资金资助项目(2009QNA5025).
摘    要:介绍文本倾向性分析的基本流程,从主观性文本识别、文本倾向性分析方法、现有系统及评测方法、语料库建设4个方面对现有文本倾向性分析技术进行介绍和概括.综述了文本倾向性分析的3类研究方法:简单统计方法、机器学习方法和细粒度情感相关性分析方法,分析这3类研究方法的特点,从算法复杂性、效率和适用范围等方面比较各自的优缺点.概括现有研究的成就和不足,从基础性问题、具体应用的实现方法2个方面提出研究的前景.


Survey on sentiment orientation analysis of texts
LI Xiao-jun,DAI Lin,SHI Han-xiao,HUANG Qi. Survey on sentiment orientation analysis of texts[J]. Journal of Zhejiang University(Engineering Science), 2011, 45(7): 1167-1174. DOI: 10.3785/j.issn.1008-973X.2011.07.005
Authors:LI Xiao-jun  DAI Lin  SHI Han-xiao  HUANG Qi
Abstract:The basic flow of sentiment orientation analysis of texts was introduced, and the primary four aspects of current interesting researches were presented: subjectivity text recognition, sentiment orientation analysis method of texts, existing systems and evaluation methods, construction of corpus. Then three methods and their characteristics were summarized, i.e. simple statistics, machine learning and fine-grained sentiment relative analysis method. Merits and demerits of methods were analyzed from complexity, efficiency and applicable scope. Finally, the current achievements and shortages were summarized, and forecasted research perspectives were proposed including basic problem and implementation method of specific application.
Keywords:
点击此处可从《浙江大学学报(工学版)》浏览原始摘要信息
点击此处可从《浙江大学学报(工学版)》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号