一种面向语义检索的向量空间模型改进方法

Improving Approach of Semantic Retrieval Oriented Vector Spatial Model

阐述了向量空间模型文本分类技术的基本思想。通过在传统的向量空间模型的基础上增加语义环境形成语义向量空间模型,用以体现领域本体概念之间的关系即语义属性。根据领域本体概念之间关系的不同层次论述了不同关系的不同权重计算方法,并且提出了单层语义向量空间模型的改进方法。

In this paper a fundamental concept about text category technology of vector spatial model was presented. The semantic vector spatial model, which could show the semantic properties, i.e. relationships between domain ontology concepts, was formed by adding the semantic environment to the traditional vector spatial model. A few weight algorithms with different relations according to the different levels of domain ontology concepts were discussed and an improved method of the single semantic vector spatial model was proposed.