摘要： 针对目前基于语义网的本体映射算法中背景本体搜索面少、本体收集不精确的问题，利用基于虚拟文档的映射技术提取在Word- Net中与概念同义的同义词集，将对单个概念进行搜索转换成对同义概念集进行搜索，从而扩大本体搜索面，获取更多背景本体。提出基于语义环境的动态本体映射算法来排除错误背景本体，使本体收集更加精确。实验结果表明，该算法可有效提高映射的查全率和查准率。

关键词: 本体, 映射, 语义网

Abstract: This paper analyzes two existing problems in ontology matching algorithm based on semantic Web. It presents a new algorithm, which uses ontology matching based on virtual document to retrieval synset in the WordNet, and changes a single concept used by ontology searching into a concept set, so it can enlarge searching scope and obtain more background ontologies. It uses dynamic ontology matching based on semantic environment to exclude wrong background ontologies, so it can get more precision in ontologies collecting. Experimental result shows precision and recall of matching improved effectively by using the improved algorithm.

Key words: ontology, mapping, semantic Web

中图分类号: 

• TP393