摘要
随着互联网的持续发展,对博客、论坛、微博等社交媒体文本的情绪分析相关研究得到学术界和工业界广泛关注。针对中文情绪分析,现有的研究主要有三类方法:基于规则和情绪词典方法、基于多类标分类机器学习方法、考虑类标间关联的方法。对几类方法进行介绍,指出存在的一些问题和发展方向。
With the continuous development of the Internet, research on the emotional analysis of social media such as blog, forum, micro-blog has been widely concerned by the academic and industrial circles. According to the Chinese emotion analysis, there are three main methods in the present study include method based on rule and expression dictionary, machine learning method based on multi label classification and method of the correlation between the emotion labels. Introduces these methods are, and points out some problems and development directions.
作者
万琪
杨祎
WAN Qi YANG Yi(College of Computer Science,Sichuan University,Chengdu 610065)
出处
《现代计算机》
2017年第2期3-5,共3页
Modern Computer
关键词
情绪分析
机器学习
多类标分类
微博
Emotion Analysis
Machine Learning
Multi Class Label Classification
Micro-Blog