摘要： 针对当前大规模本体映射分块结果不精确的问题，提出一种改进的大规模本体分块与映射方法。对本体进行预处理，把本体表示为有向无环图，将本体分块问题转换为图分割问题。采用基于遗传算法的GPO算法对有向无环图进行分块，利用基于本体块结构和参考点相结合的策略找到正确的块映射。实验结果表明，改进映射方法的分块与映射质量较优。

关键词: 大规模本体, 图分割, 遗传算法, 本体分块, 块映射

Abstract: Due to the shortcomings in current large scale ontology mapping, this paper proposes an improved method of large scale ontology partition and mapping. The ontologies are represented in Directed Acyclic Graph(DAG) structures by preprocessing. So, the ontology partition question is changed into a graph partition question. It uses Genetic Partition Ontology(GPO) algorithm which based on genetic algorithm to partition the DAG. Blocks from different ontology are matched by combining two methods of ontology blocks structure as well as predifind anchors. Experimental results show that the improved method can acquire good matching results.

Key words: large scale ontology, graph partition, genetic algorithm, ontology partition, block mapping

中图分类号: 

• TP311