模糊本体结构及基于NGD的隶属度确定

Fuzz Ontology Structure and Its Membership Determination Based on NGD

摘要:

本体能够形式化地描述特定领域的概念、术语及关系，但不能表达模糊的信息.为此将模糊描述逻辑应用在本体构建中，提出一个5元组的模糊本体模型，由概念集、角色集、实例集、模糊断言集和模糊关系集构成.基于Tableaux算法思想并结合算例，给出模糊本体模型推理问题的求解方法.针对模糊本体模型的隶属度计算问题，结合Google搜索，利用标准化谷歌距离（normalized Google distance, NGD）演算法，以实时且完全在线处理的方式计算关键词的相关性，并最终转化为模糊本体中的隶属度.

Abstract:

Ontology can formally describe concepts, terms and relationship of the special domain, but it can not express the fuzzy information. To solve the problem of the fuzzy information’s description in ontology, fuzzy description logic was employed in the construction of ontology. A 5 tuple fuzzy ontology model was established, including roles set, examples set, fuzzy assertion set and fuzzy relationship set. Based on Tableaus algorithm idea and case studies, the paper gives the solution method of the reasoning problem of fuzzy ontology. The correlation between key words was processed with the new NGD algorithm by means of Google search, and convert the results into membership degree of fuzzy ontology at last.