为了解决语义网环境下构件查询的查全查准率问题，本文提出了基于本体相似度构件查询算法。首先介绍了本体的形式化定义和构件的刻面描述方法，用本体语言（OWL）对构件本体进行形式化的描述；然后针对本体的概念相似度理论和构件刻面匹配度公式存在的不足，用构件本体的概念相似度代替公式中的匹配度，对构件刻面匹配度公式进行改进和分析，提出了本体相似度构件查询算法；算法通过计算查询树和构件树各子刻面本体概念的RDF描述相似度和RDF图结构相似度，再由构件本体相似度匹配公式进行相似度的比较，检索出相似度高的构件。实验结果表明，该算法具有理想的查全查准率。

In order to solve the problem of recall ratio and precision of the component query in the Semantic Web environment,the software component query algorithm based on ontology similarity is proposed.The paper first introduces the formal ontology definition and a component facet description method.The component ontology is formally described in the Web ontology language (OWL). Then,according to the concept of ontology similarity theory and the deficiency of the matching formula of component facet,the match degree of the matching formula of component facet is replaced by the concept similarity of component ontology;improvement and analysis are made in the matching formula of component facet.Thus,the ontology similarity query algorithm is put forward.By calculating the faceted ontological concept similarity of the component query tree and the trees corresponding to the various subcomponents,and then comparing the degree of similarity with the matching formula of component facet,the algorithm retrieves the highest degree of similarity of components.The result of the experiment indicates the algorithm has an ideal recall ratio and precision.