模糊面向对象数据模型的描述逻辑表示与推理

摘要

通过分析描述逻辑与面向对象数据模型之间的关系,研究了基于描述逻辑的模糊面向对象数据(fuzzy object-oriented data,简称FOOD)模型的表示与推理.首先,进一步研究了FOOD模型,提出了FOOD模型的形式化定义和语义定义,以便更好地在FOOD模型与模糊描述逻辑之间建立对应关系；然后,针对FOOD模型的特点和推理需求,介绍了模糊描述逻辑f-ALCIQ.在此基础上,研究了基于f-ALCIQ的FOOD模型的表示与推理,包括:实现了从FOOD模型到f-ALCIQ知识库在结构层和实例层上的转化,即实现了从FOOD模型到f-ALCIQ TBox的转化,以及从FOOD模型相应的数据库实例到f-ALCIQ ABox的转化；进而,基于转化得到的f-ALCIQ知识库,研究了如何利用f-ALCIQ的推理机制对FOOD模型的推理问题(一致性、包含性和冗余性等)进行推理；最后,设计并实现了基于f-ALCIQ的模糊描述逻辑推理机(FRsQ推理机),实现了对FOOD模型和f-ALCIQ模糊概念知识的自动推理.%The relationships between description logics and object-oriented data models are analyzed, and the paper aims at investigating the representation and reasoning of fuzzy object-oriented data (FOOD) models with description logics. The FOOD models are investigated, and the formal definition and semantics of FOOD models are proposed first. Then, aiming at the characteristics and reasoning requirement of FOOD models, the fuzzy description logic f-ALCIQ is recalled. In particular, the considers the FOOD model and the corresponding database instances simultaneously, and translate them into f-ALCIQ knowledge base at both terminological (Tbox) and assertional (Abox) levels, respectively. Furthermore, based on the translated f-ALCIQ knowledge bases, the reasoning tasks of FOOD models (e.g., consistency, subsumption, and redundancy) may be reasoned through the reasoning mechanism of f-ALCIQ. Finally, a fuzzy description logic reasonerd based on f-ALCIQ called FRsQ is designed and implemented, so as the reasoning problems above may be reasoned automatically.