摘要
在互联网环境下,不同数据源之间的语义异构问题,是制约来自多个数据源数据有效集成的主要挑战.现有语义匹配方法的不确定性,导致不同数据源之间的数据集成难以有效解决.为此,本文利用本体名称和本体结构信息这2个最重要的特征,在避免为用户带来过多负担的前提下,将用户引入传统的语义匹配过程,提出了基于用户反馈的语义匹配方法UFMatcher.实验证明,本文的方法可以在不为用户带来过重负担的前提下,有效提升匹配结果的准确率.
Under the circumstance of internet, data semantic heterogeneity constitutes the primary challenge of data integration. And the uncertainty of present automatic semantic matchers results in the problem of data integration from various data sources. In this paper we provide semantic matcher called UF-Matcher by using the name and structure, the most important features of ontology, which is based on user feedback to get the semantic relationship. Our experiments show that the UF-Matcher can settle down the problem of the uncertainty and at the same time raise the accuracy of semantic matching and reduce user burden.
出处
《北方工业大学学报》
2015年第3期49-56,85,共9页
Journal of North China University of Technology
关键词
用户反馈
不确定性
语义匹配
user feedback
uncertainty
semantic matching