基于控制词集的中文信息动态自动聚类研究  

Research on Chinese Document Dynamic Clustering Under Controlled Vocabularies

作　　者：马张华[1] 陈文广[1] 金海燕[1] 朱剑俊[1] 崔会娟[1] 刘贞君[1] 

机构地区：[1]北京大学信息管理系,北京100871

出　　处：《大学图书馆学报》2006年第6期54-60,共7页Journal of Academic Libraries

摘　　要：以专用词典为切分工具,建立以概念为基础的、具有主题分类特点的类目结构是中文信息动态自动聚类的一种适用方式。该文探索了基于控制词集的中文信息动态自动聚类技术,包括专用控制词集的构建,动态有限环境下的自动聚类程序、聚类算法,以及结合控制词集对聚类结果进行优化控制等,最后对实验结果给出了概要评价。One applicable way for Chinese Document dynamic clustering, is to use special dictionaries as segmenting tools and establish a structure with a conceptual base as well as topic category style. This paper, which explores Chinese Document dynamic clustering technology, involves the construction of special controlled vocabularies and discusses topics such as the clustering procedures, the clustering algorithms and the refinement of clustering results with controlled vocabularies, all of which are under a dynamically limited condition. In the end, the results of experiment are evaluated.

关 键 词：动态自动聚类 中文自动分类 专用切分词典 词汇控制 

分 类 号：G356.7[文化科学—情报学]