文本情感研究综述

Survey on Research of Sentiment Analysis

文本情感分类是指通过挖掘和分析文本中的观点、意见和看法等主观信息,对文本的情感倾向作出类别判断。阐述情感分析的研究应用背景,并给出基本的研究思路;分析整体的研究现状,详细回顾了主要的处理方法;在特征标注阶段,重点介绍和讨论了两类主流的处理思路———监督学习和非监督学习;简要介绍了其他一些相关的情感分析问题;总结了情感分析的现有成就和不足,以及面临的挑战,并对其发展前景进行了展望。

Sentiment analysis focuses on mining people＇s opinions and sentiments expressed in text. This paper first intro- duces the research background of sentiment analysis, and presents some basic technologies used in sentiment analysis. It then analyses the status of the whole research, and reviews the main handling in detail. Next, it focuses on an important task in sentiment analysis, opinion classification, and discusses two leading feature extraction techniques for opinion classification, supervised learning based and unsupervised learning based methods. Furthermore, it also introduces several other relevant sentiment analysis problems. Finally, the paper summarizes the current status, remaining challenges, and future directions in the field of sentiment analysis.