中文摘要：

针对语义物联网本体不利于用户理解、涵盖的内容过于宽泛等问题,提出了一种将结构特征与语用特征相结合的摘要方法。该方法以RDF语句作为本体摘要的基本单元,通过创建一个二分图模型来表征本体。每个RDF语句的重要度是它的结构重要度与语用重要度的线性总和。在重排序算法下抽取重要度高的RDF语句形成摘要,用户可以设定摘要长度。通过建立本体摘要原型系统进行实验,实验结果表明,该摘要方法能够为用户提供满足其偏好的最优摘要本体,利于用户理解本体,不但方便了语义物联网语义协同过程中对本体的相关操作,而且可以提高基于本体的信息检索的查准率。

英文摘要：

According to the SWo T ontology is not easy for users to understand and it covers too much content,to solve this problem,this paper proposed a SWo T ontology summarization method based on structural and pragmatics features. This method proposed a notion of RDF sentence as the basic unit of ontology summarization,general bipartite model and expended bipartite model of a ontology were proposed to characterize the structure in a ontology. The importance of each RDF sentence was a linear summation of both structural and pragmatics importance,summaries were produced by extracting important RDF sentence in ontology under a re-ranking strategy,a user may provide the expected length of the final summary. It designed and implemented a prototype system of SWo T ontology summarization. Some experimental result analyses show that this method can provide users with the best ontology summarization to meet their preferences,and more beneficial to ontology understanding,hence it can be more convenient for ontology-related operations in SWo T semantic collaborative,and improve the precision of ontology based information retrieval.