基于已有映射结果的本体映射  

Ontology mapping based on existing mapping result

作　　者：仲茜[1] 李涓子[1] 李毅[1] 唐杰[1] 周立柱[1] 

机构地区：[1]清华大学计算机科学与技术系,北京100084

出　　处：《清华大学学报（自然科学版）》2008年第7期1178-1181,共4页Journal of Tsinghua University(Science and Technology)

基　　金：国家自然科学基金资助项目(90604025);国家"九七三"重点基础研究项目(2007CB310803);国家"九七三"重点基础研究项目(2003CB317007);中国博士后科学基金资助项目(20070410061)

摘　　要：为了提高本体映射的查准率和查全率,提出基于已有映射结果的本体映射算法。该算法以已有映射结果为背景知识,应用多种启发式策略选择中间本体,将直接映射结果与通过中间本体的间接映射结果相结合,并利用迭代优化算法克服背景知识不准确对映射结果的影响。实验结果表明:该方法比直接映射方法的查准率和查全率分别提高1.10%和2.52%。说明该方法可以有效地提高映射的查准率和查全率。An ontology mapping method based on an existing mapping result was developed to improve both the precision and the recall of ontology maps, The system uses the previous mapping results as its background knowledge and then chooses a medium ontology by means of several heuristic approaches. The indirect mapping result from the medium ontology is then combined with the directed mapping results with optimization to revise the background knowledge for better results, Tests show that the method achieves 1.10% higher precision and 2.52% higher recall than the previous method.

关 键 词：软件工程 本体 本体映射 语义WEB 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]