文档介绍：第25卷第6期中文信息学报 , 2011年11月 JoURNAL OF CHINESE INFORMATl0N PROCESSING NoV.,2011 文章编号:1003一0077(2011)06一0118一09 文本情感倾向分析黄萱菁,张奇,吴苑斌(复旦大学计算机科学技术学院,上海201203) 摘要:近年来,文本情感倾向研究受到研究界和企业界越来越多的关注,成为了自然语言处理、信息检索、数据挖掘等领域的研究热点之一。、新问题也不断涌现。该文重点对文本情感倾向研究的前沿进展进行概括和分析。首先,结合近年来的研究成果,对文本情感倾向分析的两类主要问题进行了定义。并归纳了不同的倾向性表示方法。接下来,对倾向性分类、倾向性信息抽取,语料库与评测以及倾向性分析应用等方面的研究现状进行介绍。最后,总结了情感倾向性分析技术并对未来的发展进行了展望。由于国内对于文本情感倾向分析的研究起步较旱,在一些问题的研究上处于论文,该文也将对此加以介绍。关键词:倾向性分析;评价挖掘;倾向性分类;综述中图分类号:TP391 文献标识码:A ASurVey on Sentiment Analysis HUANG Xuanjing,ZHANG Qi,WU Yuanbin (Fudan University,School puter Science,Shanghai 201203,China) Abstmct:In recent years,sentiment analysis(opinion mining)has received increasing attentions and es one of themost popular research topics innatural language processing,information retrievaland datamining. Along with theindepth researches,a large number ofnew problems have been discovered and avariety ofnoveI sentiment analy— sismethods have been proposed. Inthis we focus on summarizing and analyzing the classical and noVel methods on thistopic. Firstly,we discuss thedefinition of thesentiment analysis and summarize the expression methods under different next sections cover thetechniques on sentiment classification,feature-based sentiment analysis,benchmark dataset, evaluation and app“cation. Finally, we conclude the topic and provid