摘要： 通过建立子类非兼类名词的上下文同现词语以及其词性的特征向量，将训练的结 果利用模糊分类方法自动构造名词子类的分类器，然后利用分类器自动识别汉语名词子类兼 类的词语，实验取得了很好的效果。

关键词: 名词, 语料库, 模糊识别, 向量空间

Abstract: The paper establishes a matrix of context vocabulary and part-of-s peech about the single-part-speech of noun-subcategory. It constructs the class ifier automatically through the method of fuzzy classification using the trainin g result. And then disambiguates the Chinese words of multi-part-speech of noun sub-category by this classifier. The experiment reaches a satisfied result.

中图分类号: 

• TP391.1