一种高效的多策略本体映射方法  

An efficient multi-strategy ontology mapping method

作　　者：姚晓明[1] 王锋[1] 林兰芬[1] 朱晓伟[1] 谢非[1] 

机构地区：[1]浙江大学计算机科学与技术学院,杭州310027

出　　处：《中国科技论文》2013年第7期642-647,共6页China Sciencepaper

基　　金：高等学校博士学科点专项科研基金资助项目(20110101110065);浙江省创新团队计划资助项目(2009R50015)

摘　　要：针对现有本体映射方法中存在的效率不高问题,提出了一种高效的基于精简本体映射概念集的多策略本体映射方法,实现了本体之间的自动化映射。通过对本体进行语言、实例、结构方面的特征分析,确定各策略使用顺序,结合层次分析法自动生成各策略结果对应的权重,最后合并各映射结果,提高映射质量。同时,在语言策略方面,充分考虑了WordNet树形层次中概念之间的语义距离与概念深度和子节点个数的关系,使语义距离计算更为准确。实验结果表明,该方法在保证查全率的同时,可有效提高查准率和映射效率。Given that existing ontology mapping methods have the problem of low efficiency, we propose a novel high-efficient multi-strategy method for automatic ontology mapping based on reduced ontology mapping concept sets to overcome this weak- ness. Based on analysis of characteristics of language, individuals and structure of the ontology, the orders of muhiple ontology mapping strategies are determined. Analytic hierarchy process （AHP） is adopted to generate the weights of each strategy, and then the results from multiple strategies are combined to improve the quality of mapping. In the meantime, the language strategy takes into account the changes of weight {ollowing the concept depth and the count of child nodes in the WordNet. Experimental results show that this method can guarantee the recall and also effectively improve the precision and efficiency of ontology map- ping.

关 键 词：人工智能 本体 本体映射 相似度 多策略 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]