文本自动分类中的动态类别扩展研究

Dynamic Class Expansion in Automatic Text Categorization

提出基于粗糙集理论的动态类别扩展算法,可以根据新文献与已有训练规则的匹配程度,有效地进行新类别的自动扩展和新分类规则的自动生成,从而屏蔽训练集和分类规则的更新等问题。

This paper proposed a dynamic category expansion algorithm in a rough set theory based text categorization system. Experiments indicated that the proposed algorithm is capable of dynamically adding new categories and classification rules as required, and organizing the incoming documents into a hierarchical classification scheme.