中文摘要：

本体作为语义web应用中语义信息的来源,在实现语义web应用的自动发现和组合中有着重要作用。当前对本体的研究多关注于提高本体使用的有效性,从而使得本体能够提供更精确丰富的语义信息,而忽略了本体自身质量对语义信息的影响。文章提出了一个基于本体概念间语义关联关系的多本体完善方法,利用参考本体中的概念关系和层次结构,结合目标本体与参考本体间的语义关联,补充与完善目标本体的概念关系和层次结构,并通过参考本体和目标本体的角色互换,迭代的完成对多本体的完善。通过对17个本体进行实验,将实验结果与传统的本体完善方法——闭包推理进行对比,证明了该方法不仅能够得到闭包推理产生的结果,并且能够对本体进行更多的正确的完善和补充,实现基于语义关联的本体完善。

英文摘要：

As the origin of the semantic information in semantic web applications, ontology is playing a cru- cial role in the implementation of the automatic discovery and composition of the semantic web applica- tions. Most current researches about ontologies focus on improving the efficiency of ontology usage in order to make the ontologies can provide more accurate and abundant semantic information. However, they ig- nore the influence of the quality of ontology on these information. This paper proposed a multi-ontology renewal method based on the semantic relations between ontology concepts to iteratively complete renewal of multi ontologies. This method uses the concept relations and hierarchy of the ontologies, combined the semantic relations between different ontologies, to supplement and improve these ontologies. Through comparing this method with the traditional renewal method （closure reasoning） based on 17 ontologies, it has proved that using this method can not only obtain the same correct renewal results as the closure rea- soning method, but also can acquire more current supplementation and renewal to realize the ontology im- provement based on the semantic relations.