基于领域本体的知识库多层次文本聚类研究——以中华烹饪文化知识库为例  

Study on Multi-level Text Clustering for Knowledge Base Based on Domain Ontology——Taking Knowledge Base of Chinese Cuisine Culture as an Example

作　　者：洪韵佳[1] 许鑫[1] 

机构地区：[1]华东师范大学商学院信息学系,上海200241

出　　处：《现代图书情报技术》2013年第12期19-26,共8页New Technology of Library and Information Service

基　　金：2011年度国家社会科学基金青年项目"联合虚拟参考咨询系统的知识库研究"(项目编号:11CTQ003)的研究成果之一

摘　　要：基于领域本体提出一种适用于知识库树状结构的多层次文本聚类方法。该方法利用领域本体将词映射为各层级的概念,先以高层级的概念实现粗粒度聚类,以识别不同题材的文本,形成知识库的主体分类框架;再结合各层级的概念与未能映射的非概念特征词实现细粒度聚类,以揭示不同深度的文本主题信息,从而实现从粗粒度到细粒度的多层次聚类。The paper puts forward a kind of multi - level text clustering method for the tree structure of knowledge base. In this method, the words are mapped as concepts by the domain Ontology. First the texts are represented by the top - level concepts to realize the big - size clustering, identify the different subjects of texts and formulate the main classification framework. Then the texts are represented by all concepts and non - concept feature words to further realize the small - size clustering and reveal the subjects of the texts with different depth. Finally, this method realizes the multi - level text clustering from big size to small size.

关 键 词：领域本体 文本聚类 知识库 中华烹饪文化 

分 类 号：G250.7[文化科学—图书馆学]