现实中的数据源一般具有半结构化、异构性和分布性等特点,而许多应用需要对不同的相关数据源进行联合操作。本体集成是解决知识共享、异构数据源语义互操作的有力工具。首先分析了本体集成的原因,提出了本体集成时应遵循的4条基本原则;然后提出了一种基于RDFS图闭包的本体集成方法,该方法将RDFS本体抽象为图模型,根据RDFS推理规则和扩展规则生成RDFS本体的图闭包,在此基础上进行本体集成,同时提出了几种计算实体间相似度的方法。最后,将该方法与FCA-merge和COMA++进行实验对比。
The data resources in real life are always semi-structure,heterogeneous and distributed.In many cases,it’s necessary to take associated operations between them.Ontology integration is a powerful tool for knowledge sharing and semantic interoperation,which comes form the different data resources.Firstly,after analyzing the reasons of ontology integration,this paper puts forward the four basic rules for ontology integration.Secondly,a method based on the closure of RDFS for ontology integration,is proposed.In this method,RDFS ontologies are abstracted as graph models.Using the RDFS inference rules and its extension,the closure of RDFS ontologies can be constructed.Based on these closures,the integration can be carried out.During this part,some measure methods for similarity between entities are proposed too.Lastly,for comparing with COMA++ and FCA-merge,this paper presents the experiments of them.