节点文献

本体复合映射发现技术研究

Personalized Service Research

【作者】 陈爱明

【导师】 徐德智;

【作者基本信息】 中南大学 , 计算机软件与理论, 2008, 硕士

【摘要】 随着语义Web技术研究的发展,本体应用越来越广泛,但是由于不同用户构建的本体在形式上、结构上的差异,导致同领域内本体在重用与交流方面存在困难,同时也限制了本体集成研究的发展,解决这些问题的关键在于本体映射的发展,虽然现有的映射方法层出不穷,但绝大部分方法却只考虑了简单的1∶1映射,没有涉及到更为复杂的复合映射问题,因此限制了映射精度的提高,本论文主要针对这个问题探索不同本体间复合映射的发现技术问题。首先,论文对课题的来源及语义Web进行了简介,由于Web自身的分布性问题,同领域内的不同用户和组织构建自己的本体,导致本体异构的产生,阻碍了本体的重用与交流。要解决异构问题关键在于映射技术,但是现有的本体映射仅仅针对了1∶1的映射问题,没有涉及复合映射问题,限制了映射精度的提高,本文针对这个问题提出了有效的复合映射发现方法。其次,根据构成复合映射的概念之间的关系的不同,提出了基于不同关系的复合映射发现技术,针对不同的关系采用不同的发现策略进行挖掘。最后,根据上述研究,本文解决了关键的技术问题、设计并实现了相关算法。并在此基础上进行了实验,也对实验结果进行了测试评估。实验结果表明,此方法具有良好的效果。

【Abstract】 With the development of semantic web techniques, ontologies have been used in more and more areas, but ontologies constructed by different users and organizations may differ from each other in form or structure, which leads to difficulty of ontology reusing and communication; meanwhile, it also blocks the research of ontology integration. The key to these problems is ontology mapping. There are many mapping methods, but most of them only focus on simple one to one mapping, while complex mapping is neglected, which results in a low mapping accuracy. This thesis is mainly aiming the problem of discovering of complex mappings between different ontologies.Firstly, the thesis introduces the sources of this thesis and Semantic web. Because of the distributing of Web, different users and organizations design their own ontologies, which leads to ontology heterogeneity, blocking the reuse and communication of ontologies. Current mapping methods only stress the one to one mapping, ignoring the problem of complex mapping, causing a low accuracy of ontology mapping results. In order to solve the problem this thesis proposed an effective complex mapping discovery method.Secondly, according to different relationships between the concepts in complex mapping, a new complex mapping method proposed based on the relationships, and different discovery technology and similarity computation methods are adopted for different relationships.Finally, based on the research above, the key technique problems are solved, related algorithms are designed and implemented; moreover, experiments are carried out and the results are evaluated. The results show that our approaches are comparatively good.

  • 【网络出版投稿人】 中南大学
  • 【网络出版年期】2009年 01期
  • 【分类号】TP393.01
  • 【被引频次】1
  • 【下载频次】84
节点文献中: 

本文链接的文献网络图示:

本文的引文网络