基于词汇相似度的元素级本体匹配

Element Level Ontology Matching Based on Lexical Similarity

随着语义Web的不断发展,本体数量日益增加。本体匹配作为本体映射、比较和集成的基础,具有重要的实际意义。由第3届国际语义Web大会(3thISWC)主办的本体匹配竞赛(EON2004)对多种本体匹配工具进行比较和评估。该文提出了一种元素级本体匹配算法LANA(LexicalAnalyzer),该算法通过计算词汇相似度得到两本体间的元素匹配对。与EON2004中的其它元素级本体匹配方法相比,LANA在没有增加匹配复杂度的基础上,具有较好的准确率和召回率。

With the development of Semantic Web, the number of ontologies is increasing greatly. In order to use those ontologies effectively, different ontologies must be mapped, compared and integrated. As the basis of these three tasks, ontology matching has great practical significance. 1SWC held a workshop called EON2004 to evaluate accuracy and recall of different ontology matching tools. This paper proposes an element level ontology matching algorithm, named LANA（Lexical Analyzer）, to map the elements of two given ontologies by calculating lexical similarity. Compared with other element level matching algorithms in EON2004 competition, LANA has higher accuracy and recall without increasing complexity of matching.