基于领域学习的本体映射方法  

Domain learning-based ontology mapping method

作　　者：利业鞑[1] 庞雄文[2] 

机构地区：[1]广东司法警官职业学院,广州510520 [2]华南师范大学计算机学院,广州510640

出　　处：《计算机应用》2009年第7期1981-1984,1995,共5页journal of Computer Applications

摘　　要：在语义数据集成中本体映射是关键,手工设置映射关系消耗时间并且不准确,需要使用本体映射工具自动发现这种映射关系。在现有本体映射方法的基础上提出了基于领域学习的映射方法,可以发现本体中概念之间的映射关系,可以从领域知识中发现复杂映射的规则、增加映射时的实例数据,提高映射发现的查全率和准确率。实验结果验证了算法的有效性。Ontology mapping is the key to semantic data integration. Discovering these mapping manually is inaccurate and time-consuming, so using tools to automatically discover these mapping between source ontology and target ontology is necessary. This paper proposed a domain learning-based ontology mapping method which can discover mapping between the concepts of ontology, used domain knowledge and historical mappings to find the complex mapping rules about the ontology being matched, and improve recall and precision. The experiment shows the correctness and efficiency of this solution.

关 键 词：本体映射 领域学习 概念依赖图 

分 类 号：TP31[自动化与计算机技术—计算机软件与理论] TP182[自动化与计算机技术—计算机科学与技术]