中文摘要：

分布式语义信息集成足语义Web面临的六大挑战之一．本体映射是语义集成的关键．文章基于贝叶斯决策理论提出最小风险的本体映射模型；RiMOM（Risk Minimization based Ontology Mapping）．RiMOM将映射发现问题转换成风险最小化问题，提供了一个多策略的本体映射方法．该方法不仅在1：1的映射上取得了较好的效果，还实现了n：1映射，实验表明在几个公开的数据集上，RiMOM可以取得比同类方法更高的查准率和查全率．

英文摘要：

Interoperability over distributed ontologies is one of the six challenges for Semantic Web. Ontology mapping is the key point to each the interoperability. In this paper, based on Bayesian decision theory, the authors propose an approach called RiMOM （Risk Minimization based Ontology Mapping） to automatically discover the mapping between ontologies. RiMOM treats the mapping problem as a decision problem and formalizes mapping discovery as that of risk minimization. Based on multiple strategies, RiMOM deals with not only 1：1 mapping, but also n：1 mapping. Experiments on several public data sets show that RiMOM can outperform existing methods in terms of precision and recall.