Font Size: a A A

Research On Multi-strategy Ontology Mapping Based On Fragments Separated

Posted on:2012-04-10Degree:MasterType:Thesis
Country:ChinaCandidate:F YinFull Text:PDF
GTID:2248330395955572Subject:Computer software and theory
Abstract/Summary:PDF Full Text Request
As a future developing direction of Internet, the goal of Semantic Web is to make the information on the Web with semantics that can be handled by computers, and to achieve the intelligent network indeed. As a method of conceptualizing domain knowledge, ontology is the key technology to realize Semantic Web. However, there often exist heterogeneities between ontologies, which become an obstacle to knowledge sharing and communicating. Correspondences between different entities form different ontologies can be discoveried by Ontology Mapping, Which is a way to solve the heterogeneous problem.The existing techniques and literature related to Ontology Mapping are deeply researched and analyzed in this paper, meanwhile, a multi-strategy ontology mapping model based on fragments separated is presented, named NSIS Model. First, for the sake of solving the problem that existing similarity calculation methods can not fully and rationally use semantic information between entities, a multi-strategy similarity calculations method—NSIS is presented from aspects of name,semantic,instance and structure, within which three single-strategy methods and way of combination are improved. Subsequently, an idea of cluster fragmentation is introduced into ontology mapping proceed. On the basis of NSIS similarity, ontology mapping method is improved and a more efficient mapping mechanism based on fragments separated is presented.Finally, experiments are designed by using the ontology sets provided by OAEI organization. By contrast, performance of mapping model presented by this paper is proved well under OAEI evaluation standard, who get better results of Precision、Recall and F-Measure.
Keywords/Search Tags:Ontology mapping, Similarity calculation, Fragments separated
PDF Full Text Request
Related items