摘要： 针对异构数据源中实现基于语义的数据集成的需求，采用分层思想，在用户层与实际数据层之间增加一个中间层来屏蔽底层数据的异构，用本体作为公共语义描述工具，建立本体到各数据源的映射规则，设计实现了将基于本体的全局查询转化为基于各数据源的局部查询系统，解决了数据源之间的语义异构问题，其中用OWL表示本体和映射，并研究了系统中的关键技术。

关键词: 语义异构, 分层, 本体, 数据集成

Abstract: Due to the demand of integration of heterogeneous data sources, a semantic-based data integration system with three layers is designed, in which a middle layer between user layer and data layer is added to hide heterogeneity of various data sources. In practice, this paper introduces the ontology used as a common semantic description, and establishes the mapping between the ontology and heterogeneous data sources, and applies Web ontology language(OWL) to describe the ontology and the mapping. A heterogeneous data integration system, which can translate ontology-based global query into data-sources-based local query, is designed and realized to resolve the problem of semantic heterogeneity. Key technologies are discussed.

Key words: semantic heterogeneity, layers, ontology, data integration

中图分类号: 

• N945