不同本体之间的异构性严重地影响了本体之间的知识共享与重用,为此,提出一种基于多种类型匹配器的本体映射方法OM-Marchers(Ontology mapping based on multiple matchers).在建立本体之间映射关系的过程中,OM-Matchers先使用多个类型的匹配器从本体模型中抽取相应类型的信息;然后这些匹配器为概念对计算相似度值,其中概念对所包含的两个概念来自于不同的本体;最后为待映射的本体模型建立相似度矩阵,并采用迭代策略完成本体映射任务.为了验证本文所提方法在处理本体映射问题时的可行性与有效性,采用OAEI所提供的共享数据集的benchmarks子集来测试OM—Matchers.实验结果表明:0M-Matchers可以有效地建立异构本体之间的映射关系.
The heterogeneity between different ontologies has seriously affected the sharing and reusing of knowledge. For this purpose, this paper presents an ontology mapping method based on multi-Matchers, called OM-matchers. During the process for creating mapping between ontologies, OM-Matchers firstly extracts the corresponding types of information, with the help of multiple marchers. Then, all the marchers calculate similarities for the concept pairs, in which concepts are from different ontologies. Finally, the similarity matrix for ontologies to be mapped is generated, and the iterative strategy is used to accomplish the work for ontology mapping. In order to verify the feasibility and effectiveness of OM-Matchers, this paper applies benchmarks, which is the subset of the shared ontology sets in OAEI, to test OM-Matchers. Experimental result shows that the method OM-Matchers can be used to create mapping between heterogeneous ontologies effectively.