基于HNC语境理论的文本分类

基于HNC语境理论的文本分类

王文峰¹  唐兴全²

北京大正语言知识处理研究院 北京 100081

                    对外经贸大学中文系 北京 100083

¹ Wangwenfeng2008@hotmail.com, ² tang_xq@hotmail.com

摘要：本文从HNC语境理论来分析现在中文信息处理的一个热门话题——中文文本分类。HNC理论认为只有深入到语境层面，理解语句及篇章，才能更好的完成文本分类的工作。本文介绍了一种新的文本语义形式化模型——语境框架。语境框架是一个三维的语义描述，他把文本内容抽象成领域（静态范畴）、情景（动态范畴）、背景（褒贬、参照等）三个框架。并在语境框架的基础上，设计实现了文本特征提取算法。

关键词：HNC 语境框架 文本分类

The Text Classifies based on the HNC theory

Wang Wenfeng¹  Tang Xingquan²

¹ Linguistry Management Institute&Com.Ltd, Dazheng, Beijing 100081

² department of Chinese University of international business and economics  Beijing 100083

 ¹ Wangwenfeng2008@hotmail.com, ² tang_xq@hotmail.com

Abstract:This article analyzes the present Chinese information processing from the HNC linguistic environment theory that is a hot topic of discussion - - Chinese text classification. The HNC theory thinks only when penetrated into the linguisticenvironmentstratification plane and understood the sentence as well the chapter ,which can complete the work of the text classification further. This article introduces a kind of new text semantics formalization model - - linguistic environment frame. The ling