BP神经网络在网页自动分类中的应用

BP Neural Network and Its Application in Web Document Automatic Classification

针对信息挖掘中的网页自动分类问题,提出了一种基于向量空间模型和并联BP网络的分类方法。该网络由并行连接的多个子网络组成,每个子网络负责一类模式特征的提取,多个子网并行处理所有模式,将分类结果在总输出层表现出来。以因特网上旅游网页分类为例验证了该方法的有效性。

Aiming to web document classification in data mining, a classification method is presented in this paper. The method is based on vector space model and parallel connection BP neural network. The model includes some parallel connecting sub- networks, each of which accounts for extracting a sort of pattern. Total subnetworks synchronously deal, with all patterns, and present classfication eafion results in last output layer. The availability of model is proved by classification of some web documents in Intemet.