一种OWL本体进化方法  

Approach of OWL ontology evolution

作　　者：高俊杰[1] 邓贵仕[1] 

机构地区：[1]大连理工大学系统工程研究所,辽宁大连116024

出　　处：《计算机应用研究》2009年第7期2564-2567,共4页Application Research of Computers

基　　金：国家自然科学基金资助项目(70671016)

摘　　要：针对从关系数据库模式学习所得的OWL本体大都是轻量级的,其概念层次结构过于扁平,很难被直接用于实际的本体应用,提出一种新颖的OWL本体进化方法。其通过形式概念分析对已有轻量级OWL本体进行概念聚类,根据概念等同度、概念包含度计算,自动提出丰富、修改本体概念语义关系的建议,从而辅助设计者实现本体进化。该方法将FCA与相似度计算结合使用,既发挥FCA语义强度较高的特点,又发挥相似度计算执行效率高且容易实现的特点;同时,规避了相似度计算语义强度较低与FCA实现较为困难且执行效率较低的不足。一个实例结果的评估证实了该方法的有效性。OWL ontologies learned from relational database schema are almost lightweight, their concept hierarchies are too horizontal, it is difficult to apply them to construct ontology-based systems directly. Aiming to this new problem, this paper proposed a novel approach for OWL ontology evolution. Based on concept clustering using formal concept analysis, computing con- cept equation measure and concept inclusion measure, made some suggestions to enrich or amend concept hierarchy of ontology automatically to aid designer to achieve ontology evolution. Validation of the approach was done by the evaluation of an experiment result.

关 键 词：关系数据库模式 本体进化 概念聚类 OWL本体 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]