一种基于RBF神经网络的XML文本分类方法

A Text Categorization Method Based on RBF Neural Network

为了快速有效地组织和分析海量的XML文本信息,XML文本的自动分类必不可少。文中提出了一种基于RBF神经网络的分类方法,并运用改进型的CHI统计量方法进行特征提取,对传统的加权公式进行了一些改进,再运用资源优化神经网络(RON)进行训练,做了必要的实验分析。实验结果表明该分离器有较高的分类质量,提高了分类的效率,有较高的分类准确性,满足了XML文本自动分类的要求。

In order to organize and analyse the mass information of the XML text messages quickly and efficiently, the automatic classification of XML text is essential. Proposes an classificstion method based on RBF neural network, and uses an improved CHI statistic methods to extract feature, improves the traditional weighted formula. The results show that the separator has a higher quality of the classificstion, to improve the categorizing effectiveness and to meet the XML text categorization.