基于混合本体的文献分类研究--以计算机学科为例

摘要

Traditional literature classification can't satisfy the requirements of literature organization and user demand. The article, taking computer science discipline as an example, makes a comparative study on traditional literature classification and Folksonomy to analyzes their advantages and disadvantages respectively, and proposes an ontology construction method combining traditional literature classification and Folksonomy to classify literature. The article studies tag extraction and semantic mapping approach based on Folkson-omy, and tries to build computer literature ontology by taking"Douban"reading as a reference, so as to verify the classification. The ex-periment shows the method can organize literature resources effectively, and can meet users demand with practical significance for liter-ature classification.%传统的文献分类方法已经无法很好地适应文献组织和用户的需求。本文以计算机学科为例，对传统文献分类和Folksonomy进行对比研究，分析两者的优缺点，提出传统分类和基于Folksonomy的混合本体构建方法，研究了基于Folksonomy的标签抽取和语义映射方法。并以豆瓣读书为参考，尝试构建计算机文献本体，验证了该分类方法可有效地组织文献资源、满足用户需求，对规范文献分类具有实际意义。