基于SUMO本体的中文文本自动分类研究  

Research on Chinese Text Automatic Classification Based on SUMO Ontology

作　　者：马芳 

机构地区：[1]烟台工程职业技术学院图书馆,山东烟台264006

出　　处：《情报科学》2015年第6期43-47,共5页Information Science

摘　　要：在文本分类中,传统的文本表示模型忽略了词语之间的内在关联性,而真实的文本中存在较复杂的语义关系。针对此问题,将SUMO本体引入文本分类模型,利用本体的概念语义关系和清晰的层次结构,对文本内容进行概念映射,使最终特征空间包含更多的语义信息。实验表明,该模型可以有效地提高文本分类的准确性,具有一定的研究意义。In text classification, the traditional text representation model ignores the inherent relationship between words, whereas in the real texts complex semantic relationship exists. Aiming at this problem, in- troduce the SUMO ontology into text classification model, using the concept of ontology semantic relations and clear hierarchy, mapping the text content to concept, so that the final feature space contain more se- mantic information. Experiments show that, the model can effectively improve the accuracy of text classifi- cation, which has certain research significance.

关 键 词：文本分类 本体 向量空间 

分 类 号：G254[文化科学—图书馆学]