Abstract: The traditional search methods suffer from the problems of high time complexity and easily getting wrong candidate mapping sets. This paper proposes an improved method of search candidate mapping sets. This method obtains the initial candidate mapping sets through comparing the similarity between the different concepts, and extends them by making use of correlation to get the final candidate mapping sets. Compared with traditional methods, experimental results prove that this method is valid.

Key words: ontology, correlation, candidate mapping sets

摘要： 针对传统的本体候选映射集搜索方法时间复杂度偏高、且容易得出错误的候选映射集等问题，提出一种改进的候选映射集搜索方法。该方法通过比较概念间的名称相似度获得初始的候选映射集，利用概念间的相关度对其进行扩展，从而得出最终的候选映射集。实验结果验证了该方法的有效性。

关键词: 本体, 相关度, 候选映射集

CLC Number: 

• TP311