Computer Engineering and Applications ›› 2010, Vol. 46 ›› Issue (23): 159-161.DOI: 10.3778/j.issn.1002-8331.2010.23.045

• 数据库、信号与信息处理 • Previous Articles     Next Articles

Dimensionality reduction of short message text classification and thematic extraction of semantic

LIU Jin-ling   

  1. Huaiyin Institute of Technology,Huai’an,Jiangsu 223003,China
  • Received:2009-01-19 Revised:2009-03-20 Online:2010-08-11 Published:2010-08-11
  • Contact: LIU Jin-ling

基于降维的短信文本语义分类及主题提取

刘金岭   

  1. 淮阴工学院 计算机工程系,江苏 淮安 223003
  • 通讯作者: 刘金岭

Abstract: In order to predict Chinese short message opinion quickly,the short message text dimensionality reduction is obtained by using synonymous terms and the upper and lower merge,and then the mass short message classification algorithm for the extraction and classification of the subject is given.Test shows that the method can greatly improve the public opinion of the speed and quality of forecasts.

Key words: classification, short message text, dimensionality reduction, theme

摘要: 为了对中文短信文本进行快速的舆情预测,利用对同义关系词汇归并和上下位词汇聚焦以及种子词汇的确定来实现对短信文本空间的降维,而后又给出了海量短信文本分类的算法及分类主题的提取。实验表明该方法可以大大提高舆情预测的速度和质量。

关键词: 分类, 短信文本, 降维, 主题

CLC Number: