中文微博情感分析方法研究  

Research on Sentiment Analysis Method of Chinese Microblogging

作　　者：王银[1] 吴新玲[1] 

机构地区：[1]广东技术师范学院计算机科学学院,广东广州510665

出　　处：《广东技术师范学院学报》2014年第3期63-68,共6页Journal of Guangdong Polytechnic Normal University

摘　　要：微博情感分析是当今情感分析的一大研究热点,中文微博情感分析按照处理层面的不同可以分为主题无关和主题相关分析,分析方法主要分为两类,一是基于情感词典的方法,另一类是基于机器学习的方法.本文对目前中文微博情感分析的主要流程和处理方法进行了探讨,对两类方法的特点及其研究现状进行了分析、比较,对不能完全准确进行情感分类的主要原因进行了归纳总结,为情感分析方法的进一步研究提供了参考方向.Chinese microblogging sentiment analysis is a research hotspot now. Chinese microblogging sentiment analysis according to different processing levels can be divided into the target-independent analysis and target-dependent analysis, analysis methods can be divided into two categories, one is the method based on emotion lexicon, another is the method based on machine learning. This paper discusses the main process and processing method of Chinese microblogging sentiment analysis at present. The characteristics of two methods and research status are analyzed and compared, the main reason for not fully accurate emotion classification are summarized, providing reference to further research of sentiment analysis method.

关 键 词：中文微博 情感分析 情感词典 机器学习 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]