A multilingual ontology-based approach to attribute correspondence identification

The paper focuses on studying a new problem in enterprise data integration, i.e. attribute correspondences identification in multilingual schemas, and proposes a multilingual ontology-based method for the problem. In this paper, the problem is analyzed through comparing two schemas of a large financial corporate of China. And based on the multilingual ontology, attribute correspondences are identified in multilingual schemas. Moreover, a computer-aided system to deal with the problem according to the method is given. A prototype has been implemented, which shows its effectiveness in practice.