【作者】 吕艳辉；

【导师】 马宗民；

【作者基本信息】 东北大学 ， 计算机应用技术， 2010， 博士

【摘要】 语义Web是当前Web的扩展,它赋予Web资源信息机器可理解的语义,从而便于人和计算机之间的交互与协作。为了让机器能够理解Web资源信息并做推理,需要建立本体,并使用本体语言进行描述。语义Web本体语言的标准是OWL,它建立在数据表示语言RDF和RDFS(常合称为RDF(S))之上,它们一起构成了当今语义Web的描述语言基础。由于很多语义Web应用需要处理大量的模糊知识,而现有本体不能直接用于模糊知识的表示和处理,因此,对本体进行模糊扩展以满足模糊知识管理的需要逐渐成为一个研究热点,这一点与数据库技术为表示和处理现实世界中的模糊数据而产生模糊数据库模型的情况相一致。作为Web时代模糊信息表示和处理的两个重要技术方法,模糊数据库模型和模糊本体之间存在着密切的关联关系。一方面,从构建的角度,模糊数据库模型可以作为构建模糊本体的数据源,使模糊本体充分利用现有的模糊数据库模型中的信息；另一方面,从存储的角度,利用模糊关系模型亦即模糊关系数据库在模糊数据存储和处理等方面的优势,能够对语义Web上的模糊信息进行有效的管理。事实上,模糊本体构建和存储是模糊本体管理中的两个重要问题。目前,有关模糊本体存储以及利用结构化模糊数据进行模糊本体构建的研究成果还很少。为有效表示和处理大量的模糊知识、实现模糊语义Web本体的管理,本文在对语义Web数据层语言RDF(S)及本体层语言OWL模糊扩展的基础上,展开数据库支持的模糊OWL本体管理的研究,目标在于形成一个有关模糊OWL本体从表示到构建、存储的完整理论框架。具体研究内容包括以下几个方面：(1)提出模糊RDFS(f-RDF(S))和模糊OWL(f-OWL)。通过对RDF数据类型进行模糊扩展,给出模糊数据类型的表示方法,这种方法能够解决模糊数据类型信息的表示和处理问题,进而,提出f-RDF(S),并给出f-RDF(S)语义。在f-RDF(S)的基础上,对OWL进行模糊扩展,提出f-OWL,并给出f-OWL的语法和语义。f-OWL奠定了描述模糊OWL本体的语言基础,在此基础上,给出模糊OWL本体的形式化定义。(2)提出基于模糊EER模型的模糊OWL本体的构建方法。在给出模糊EER模型的形式化定义后,从模糊描述逻辑角度,指出基于模糊EER模型构建模糊OWL本体具有可行性。在此基础上,给出基于模糊EER模型构建模糊OWL本体的步骤,提出模糊EER模型到模糊OWL本体的转换规则和算法。(3)提出基于模糊关系数据库的模糊OWL本体的构建方法。在给出模糊关系数据库的形式化定义后,从模糊描述逻辑角度,指出基于模糊关系数据库构建模糊OWL本体具有可行性。在此基础上,给出基于模糊关系数据库构建模糊OWL本体的步骤,包括模糊关系模式的语义识别、转换规则和算法,以及利用模糊关系数据库中的数据构建模糊OWL本体实例的方法,并证明了基于模糊关系数据库的模糊OWL本体构建方法的正确性。(4)提出基于模糊关系数据库的模糊OWL本体的存储方法。通过分析现有本体的存储方法,给出模糊OWL本体的存储模式,该存储模式能够满足模糊OWL本体的存储要求。在此基础上,提出模糊OWL本体结构和不同类型的模糊数据在模糊关系数据库中的存储方法,实现了模糊OWL本体在模糊关系数据库中的合理、有效的存储,并证明了基于模糊关系数据库的模糊OWL本体存储方法的正确性。更多还原

【Abstract】 The Semantic Web is an extension of the current web in which the web information can be given well-defined semantic meaning, and thus enabling better cooperation between computers and people. In order to recognize the web resources by computers in an intelligent and automatic way, we should construct ontologies and use ontology description languages to represent ontologies. OWL is a standard for expressing ontologies in the Semantic Web building on data description language RDF and RDFS (or simply RDF(S)). OWL together with RDF(S) forms the foundation of current Semantic Web description languages.Classical ontologies have limitations when dealing with fuzzy knowledge and data information that play an important role in many web applications, so the significant research efforts in the Semantic Web community are recently directed toward the fuzzy extensions to ontologies, which is in accordance with the case in databases that the fuzzy database models are developed for representing and processing the fuzzy data in many real-world applications.As two important techniques of representing and processing the fuzzy information in web, there exist strong relationships between the fuzzy database models and the fuzzy ontologies. On the one hand, from the view of construction, the fuzzy database models in the database area can be used as the dominant source for acquiring the existing knowledge for fuzzy ontologies development. On the other hand, from the view of storage, the fuzzy information in Semantic Web can be managed more effectively by taking advantage of storing and processing fuzzy data of fuzzy relational models (i.e., fuzzy relational databases). In fact, the construction and the storage of fuzzy ontologies are two significant issues in fuzzy ontologies management. Currently, less research has been done in the storage of fuzzy ontologies and the construction of fuzzy ontologies from the structured fuzzy data.In order to represent and deal with large amounts of fuzzy knowledge, and realize the management of fuzzy Semantic Web ontologies, the fuzzy extensions of the data level and the ontology level languages of Semantic Web are firstly investigated. On this basis, the management of fuzzy OWL ontologies supported by databases is investigated in detail so as to form a comprehensive theoretical framework of representation, construction, and storage of fuzzy OWL ontologies. The main contributions of this paper include:(1) The fuzzy RDFS(f-RDF(S)) and fuzzy OWL (f-OWL) are proposed. Firstly, based on the mechanism of RDF data type, a method of representing the fuzzy data type is proposed, which can solve the problems of representing and processing the fuzzy data type information. The f-RDF(S) is hereby presented, and the semantics of f-RDF(S) is also given. Then, based on f-RDF(S), the OWL is extended to f-OWL, and the syntax and semantics of f-OWL are given. The f-OWL lays the language foundation of describing the fuzzy OWL ontologies. On this basis, the formal definition of fuzzy OWL ontologies is given.(2) The construction method of fuzzy OWL ontologies based on fuzzy EER models is presented. Firstly, the formal definition of fuzzy EER models is given, and then from the view of fuzzy description logics, the feasibility of constructing fuzzy OWL ontologies from fuzzy EER models is discussed. On this basis, the steps of constructing fuzzy OWL ontologies using fuzzy EER models is given, and the transformation rules and the algorithm of fuzzy EER models to fuzzy OWL ontologies are given.(3) The construction method of fuzzy OWL ontologies based on fuzzy relational databases is presented. Firstly, the formal definition of fuzzy relational databases is given, and then from the view of fuzzy description logics, the feasibility of constructing fuzzy OWL ontologies from fuzzy relational databases is discussed. On this basis, the steps of constructing the fuzzy OWL ontologies using the fuzzy relational databases, including the semantic recognition of fuzzy relational models, the transformation rules, and the algorithm are given. Then, the generation of fuzzy OWL ontologies instances using the data of fuzzy relational databases is developed in detail. The correctness of construction method of fuzzy OWL ontologies is proved.(4) The storage method of fuzzy OWL ontologies in fuzzy relational databases is proposed. Firstly, by analyzing the storage approaches of ontologies, the storage schema of fuzzy OWL ontologies is given, which can satisfy the storage demand of fuzzy OWL ontologies. On this basis, the storage methods of fuzzy OWL ontologies structure and fuzzy data of different types in the fuzzy relational databases are proposed, which achieve reasonable and efficient storage. The correctness of storage method of fuzzy OWL ontologies is proved.更多还原