摘要
为了解决本体映射方法中计算量大、方法单一的问题,提出本体相似度综合映射方法。首先分解合适的本体,将规模比较大的本体分解为小本体,以降低映射计算的时间复杂度;然后根据本体映射的启发规则筛选出候选概念集,对候选概念集进行基于世界知识体系的本体概念相似度计算,再进行语义相似度和结构相似度计算,并把这3种不同算法得到的相似度值进行加权综合,给出最终的本体概念相似度值,再根据该值进行本体映射;最后通过设计实验来验证算法的正确性与有效性,结果表明本方法能在提高映射效率的同时保证良好的查询效果。
Aiming to the problems of large amount of calculation and single method on the calculation of the concept in ontology mapping, this paper presents a comprehensive mapping method, that decomposes with suitable ontology at first, making the large size of ontology into small ontologies, reducing the computational time complexity of mapping;Then the candidate set of concepts are selected according to the ontology mapping heuristic rule, calculating the similarity of ontology concepts based on the HowNet, then computing the semantic similarity and structure similarity, the next to do is using the comprehensive weight to get the final comprehensive ontology concept similarity;Finally, an experiment is made to get the validity and correctness of algorithm. The results show that method can well improve mapping efficiency and effect of the mapping method is pretty good.
出处
《计算机与现代化》
2014年第11期61-65,共5页
Computer and Modernization
关键词
异构本体
本体分解
本体映射
相似度计算
时间复杂度
heterogeneous ontology
ontology decomposition
ontology mapping
similarity calculation
time complexity
