【作者】 李明泉；

【导师】 冯志勇；

【作者基本信息】 天津大学 ， 计算机应用技术， 2008， 博士

【摘要】 计算机与Internet的快速发展,使得语义网和描述逻辑在人工智能等领域扮演着越来越重要的角色,基于语义网的知识获取与知识推理应用也在不断增加。但在我们的日常生活中,越来越多不确定的、模糊的信息需要处理,这就需要有一套完整的理论联系实际的模糊知识处理体系,来对模糊信息进行准确的表述与推理。本文以此目标为研究方向,把模糊逻辑理论作为基础,对现有的基于语义网的理论与技术进行了深入的研究,重点关注了理论的完善和实践的可应用性两个关键的问题,所取得的主要研究成果如下:(1)提出了对描述逻辑SHIQ的模糊扩展,给出了一个可靠的公理体系:模糊描述逻辑公理体系F-SHIQ,将其作为知识表示与推理的理论基础。以一个构造公理系统的角度,将描述逻辑中必要的概念、关系、操作等都进行了模糊扩展,给出了相应的定义,法则,公理和定理等,并对公理体系进行了推导证明,以保证其可靠性。(2)把F-SHIQ公理体系作为逻辑基础,扩展了现有的本体描述语言OWL DL,提出了基于模糊描述逻辑的规则语言FSRL。使得知识表述语言在类、属性、个体、操作等方面同样具有了表达和处理模糊信息的能力。并为后面的模糊推理机提供了必要的知识表示形式基础,推理机中的知识和规则都可以以此形式进行存储应用。(3)提出了使用本体和语义规则来对模糊公理体系的法则,定理等进行存储的存储模型。利用本体是个语义知识库的特点,把公理体系分解为最基本的概念类和概念关系,并利用语义规则进行必要的关联。这样既高效的实现了对法则定理等的全面存储,又消除了存储的冗余。(4)以F-SHIQ为逻辑基础,以FSRL为知识表示与存储形式,给出了一个能够推理模糊知识的实用型模糊推理机FSRLReasoner。对以知识基形式存在的模糊问题,解析后利用规则进行模糊推理,给出相应的结果,最后还通过一个知识基推理的实验,对本推理机的能力进行了必要的检验。本文的研究工作,将模糊逻辑引入扩展到了语义网的知识表述与推理中,增强了处理模糊信息的能力,提供了一整套的模糊信息处理体系,对今后各学科模糊知识的表述与推理都有着很好的理论和实践意义。更多还原

【Abstract】 With the rapid development of computer and Internet, semantic web and description logic become more and more important in artificial intelligence and other fields, and the knowledge acquisition and knowledge reasoning based on semantic web has also increased. However, in our daily life, more and more uncertain and incomplete information needed to be resolved, so this requires a complete system with theory and practice to deal with the vague knowledge, including the representation and reasoning of fuzzy information. In this paper, we make the fuzzy logic as the basis theory, learn existing theories and technology based on the Semantic Web, and focus on the improvement of the theoretical and practical applicability of the two key issues. The main research results are as follows:(1)An Fuzzy Description Logic Axiom System named F-SHIQ is proposed as basic theroty for knowledge representation and reasoning. And a fuzzy extension of necessary concept, relations and operation of description logic is given in the point of axiom system construction. This axiom system gives corresponding definitions, rules, axioms, theorems and reasoning proofs of theorems for the dependability.(2)A fuzzy SHIQ rule language FSRL based on F-SHIQ axiom system is proposed to extend existing ontology language OWL DL. The extension makes the knowledge expression language having the capability of fuzzy information representation and reasoning in the classes, properties, individuasl, and operations. And FSRL provides the necessary forms basis of knowledge expressions for the fuzzy reasoning behind. The knowledge and rules in reasoner are saved for reasoning application by FSRL.(3)A storage model using ontology and semantic rules is proposed to save the algorithms and theorems of fuzzy axiom system. The model broke down axiom system into simple concepts and relations, use the ontology as a Knowledge DataBase to save and build link to them via semantic rules. In this way, an efficient storage of the algorithms and theorems is realized to eliminate the redundancy.(4)Using F-SHIQ as the logic basis and FSRL for the knowledge expression and storage form, a fuzzy engine FSRLReasoner is given to reasoning fuzzy knowledge. For the fuzzy issue of Knowledge Base form, the reasoner parses it and then does fuzzy reasoning by the rules, gives the corresponding results. Finally, an experiment of KB reasoning is applied to prove the capacity of this reasoner.The research introduces the fuzzy logic to represent and reason knowledge in semantic web, and enhance the ability of dealing with fuzzy information. It also provides a set of fuzzy information processing system, and has a well theoretical and practical significance to the various field in the future fuzzy knowledge dealing.更多还原