基于类别空间模型的文本倾向性分类方法  

Text tendency categorization method based on class space model

作　　者：李艳玲[1] 戴冠中[1] 朱烨行[1] 

机构地区：[1]西北工业大学自动化学院

出　　处：《计算机应用》2007年第9期2194-2196,共3页journal of Computer Applications

摘　　要：在对现有分类方法和文本倾向性分类的复杂性进行分析的基础上,提出了一种基于类别空间模型的文本倾向性分类方法。该方法采用组合特征提取方法,基于词语对类别的倾向性进行分类。实验结果表明该方法有效地提高了倾向性分类的精度和速度。Text tendency categorization has important application in the content security management and pubhc oplmon analysis. Based on the studies of existing classification methods and the complexity of text tendency categorization, a text tendency categorization method based on class space model was proposed. By means of combined feature selection method, the new method realized text tendency classification based on the words of the types of tendency. Experimental results show that the method is effective in improving the accuracy and speed of text tendency categorization.

关 键 词：文本倾向性分类 类别空间模型 特征提取 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]