Computer Engineering and Applications ›› 2007, Vol. 43 ›› Issue (4): 167-169.

• 数据库与信息处理 • Previous Articles     Next Articles

Research on Text Genre Classification Based on Sentiment of Features

YanHui Zhu   

  • Received:2006-07-07 Revised:1900-01-01 Online:2007-02-01 Published:2007-02-01
  • Contact: YanHui Zhu

基于特征情感色彩的文档流派分类研究

朱艳辉 阳爱民 王平   

  1. 株洲工学院 重庆邮电学院自动化研究所
  • 通讯作者: 朱艳辉

Abstract: The most difference between genre classification and text classification which based on topic is the text feature. Genre classification requires that text feature should contain more semantic information especially information on document style. Classification based on sentiment of features appends semantic information such as sentiment to text feature.This paper introduces conception and applications of text genre classification firstly, Then analyzes the text feature of genre classification and several algorithms to get word sentiment orientation weight., discusses the processes of text genre classification based on sentiment of features. Experiment results analyse and comparison are given at last.

Key words: genre classification, text feature, sentiment, weight

摘要: 流派分类和基于主题的文本分类最大的区别之处就在于文本的特征。流派分类需要能够描述文档风格的、表达更强语义信息的特征,基于特征情感色彩的分类方法是将情感色彩这种语义信息附加到特征上。文章首先介绍了文档流派分类的概念及其应用,然后分析了流派分类的文本特征和词汇的情感倾向权值的几种计算方法,论述了基于特征情感色彩的文档流派分类过程,最后对几种分类方法进行了实验结果分析和比较。

关键词: 流派分类, 文本特征, 情感, 权值