【摘要】 提出了一种基于文档集重心向量分类思想的概念特征提取方法。该方法从文档集中提取表述概念特征的向量,挖掘概念特征深层次的差异,表达文档中蕴含的抽象概念,为基于领域知识库的概念网构建提供良好的先决条件。更多还原

【Abstract】 A method for extracting concept features of centroid-based document classification algorithms is presented. By extracting vectors whichpresent concept features from documents, the deep-seated differences of concept features are discovered and the abstract concept informationcontained in the documents is expressed, which in turn, is the presumption of concept network construction based on domain knowledge base.更多还原