面向多本体语义相似度计算的改进研究

Improvement of Semantic Similarity Computation Based on Multiple Ontologies

相对于仅利用单一本体度量语义相似度,多本体所蕴含的额外信息可提升语义相似度计算的精准度。通过对现有多本体语义相似度计算方法的分析,提出充分利用多本体中匹配概念对所提供线索的改进相似度计算方法。实验结果表明,改进的方法能够在一定程度上提升语义相似度度量的准确度。

Comparing to measure the semantic similarity with a single ontolgoy, the additional information contained in the muhiple ontologies can be used to improve the accuracy of semantic similarity computation. After analyzing the recent computation method of semantic similarity based on multiple ontologies, proposes an improved computation method by making full use of the matching concept pairs in the multiple ontologies. The experimental result presents the improved method can enhance the accuracy of semantic similarity to a certain extent.