摘要
分析了微博文本的特点,设计了微博信息监测应用的功能。采用VSM模型表示微博文本,并基于SVM分类算法对微博评论进行情感分类;结合评论者的身份信息对其权威性进行计算。最后综合每位评论者的情感倾向及其权威性,得到大众对该微博话题的情感倾向。
The characteristics of micro-blog texts are analyzed, and the fimctions of a micro-blog information monitoring system are designed. Then the VSM model is used to represent the text of the micro-biog. The comments in micro-blogs are classified according to emotion based on SVM classification algorithm and the authority of every commenter is calculated using his personal information. Combining the emotional tendency and authority of each commenter, the public emotional tendency to this topic may be estimated.
出处
《计算机时代》
2012年第6期7-9,共3页
Computer Era
基金
浙江省社科联研究课题(2011B097)
关键词
情感分类
微博
SVM
特征选择
sentiment classification
micro-blog
SVM
feature selection