【作者】 欧阳杨；

【导师】 朱淼良；

【作者基本信息】 浙江大学 ， 计算机应用技术， 2008， 博士

【摘要】 随着信息爆炸式的不断增长,不同学科领域知识的多样性和异构性使知识的表达和共享成为难点。同时,自主学习模式受到越来越多学习者的青睐,如何在复杂的知识库中获取到所需的知识并构成自己的知识系统成为研究的热点。教育语义网通过将语义网技术与网络教育技术相结合,为现代教育提供了新的发展契机,旨在为学习者提供更加高效高质量的智能服务。本文以教育语义网为背景,用本体技术对领域知识和用户模型进行语义化描述,提出基于本体的自适应学习系统构架(OntoAES,Ontology-based Adaptive E-learning System),提供知识的获取和共享平台,为学习者提供基于个人知识空间和偏好的学习服务。本文首先研究了教学和学习过程中的各种理论模型,分析理论模型中对学习行为的定义和特征描述,基于各种学习行为的特征,研究自适应学习过程的特点和需求,为设计自适应学习系统提供理论框架和行为模型。研究了知识空间的表达方法,构建了OntoAES领域知识模型和用户知识空间模型。其次,研究了教育领域知识的特点,针对领域知识的复杂性和多样性以及领域知识专家对本体工程技术的缺乏造成的本体开发难度加大的问题,提出了基于知识工程的OntoAES本体构建方法,提出了领域知识概念的提取、层次结构的定义以及关系模型的建立过程与方法,简化了学科领域本体建模过程。再次,基于领域知识空间模型,以计算机学科领域为研究对象提出了计算机学科领域的知识分类体系和概念集合,提出了OntoAES计算机学科领域知识本体的构建方法。在学习资源标准的基础上建立了学习资源描述本体,赋予了学习资源描述模型更加丰富的语义和更大的可扩展空间。其四,研究了用户信息模型与用户知识空间模型,建立了OntoAES用户模型本体,分别构建了描述用户基本信息的用户信息子本体,描述用户偏好信息的用户偏好子本体,描述用户绩效信息的用户绩效子本体,构建了描述学习技能的学习能力本体。分析了用户模型本体与领域知识本体和学习资源本体间的语义关联,为自适应系统的构建建立了良好的模型基础。最后,提出了基于本体的自适应学习系统OntoAES的功能模块与系统构架,研究了各个功能模块与本体间的关联和应用模式,研究了基于本体的自适应学习过程,详细分析了OntoAES系统中自适应学习的步骤和系统流程。通过对用户学习行为的研究,基于用户使用日志对潜在的学习资源关系模式进行了详细的分析和定义,基于信息提取和数据挖掘技术,建立资源的关系模型和用户偏好模型,使得在自适应学习系统中可以自动的更新领域知识学习路径信息和用户偏好信息,验证了本文所提出的模型和方法。本文的创新点主要体现在:1)提出了知识空间的关系模型,提出了基于知识工程的领域知识本体构建方法;2)研究了计算机科学领域本体和用户模型本体,提出了基于本体的自适应学习系统框架;3)定义了基于本体的自适应学习规则,建立了用户偏好本体与学习资源描述本体的匹配模型和基于使用数据的学习资源关系模型和偏好模型。本文的研究工作为教育语义网中的本体构建和自适应学习系统的设计和构建提供了研究基础和指导,今后的研究方向包括本体的补充和完善及自适应学习系统功能的完善,研究本体间的匹配模型以及基于本体的自适应学习规则,为学习者提供智能高效的学习系统等。更多还原

【Abstract】 As information increases explosively, the diversity and heterogeneity of knowledge in different domains make it difficult to represent and share knowledge. Meanwhile, the adaptive learning pattern is favored by more and more learners, and how to acquire the knowledge in need from complex knowledgebases and construct personal knowledge system has recently become a hot spot of research. Semantic learning web, which combines semantic web and web-based education technologies, shed lights on the development of modem education, and provide learners more efficient and high-quality intelligent services. This paper, which based on semantic learning web, semantically described domain knowledge and user pattern using Ontology technology, presented the architecture of ontology-based adaptive e-learning system (OntoAES) , provided the platform for knowledge acquiring and sharing, and also provided learners with effective learning services based on personal knowledge spaces and preferences.Firstly, various theory models of teaching and learning processes were studied; the definition and description of learning behaviors in those theory models were analyzed; based on different characteristics of the learning behaviors, the features and requirements of the adaptive e-learning process was studied in order to provide the theory architecture and behavior model for the adaptive e-learning system; how to present knowledge space was studied, domain knowledge model and user knowledge space model were established.Secondly, the features of domain knowledge was studied; as the complexity and diversity of domain knowledge and the lack of ontology engineering technology for domain experts make it difficult to develop domain ontology, the method to establish ontology based on knowledge engineering was proposed; the method to extract domain knowledge concepts, define concepts hierarchical structure and construct the relationship models were presented. The construction process to build domain ontology was simplified. Thirdly, computer science was choosen as the research domain, based on the domain knowledge space model, the knowledge taxonomy architecture and concept sets were constructed, and domain knowledge ontology was built. On the basis of E-learning standards, learning resources description Ontology was established, which provided more semantics to learning resources description model and more space to be expanded.Fourthly, the user information model and user knowledge space model were studied, the user model Ontology was built. The user information sub-Ontology to describe user’s basic information, the user preference sub-Ontology to describe user’s preference information, user performance sub-Ontology to describe user’s performance information along with the user competency Ontology to describe user’s learning skills were established respectively. The semantic association among user model ontology, domain knowledge ontology and learning resources description Ontology was analyzed. A well-modeled basis was build for the adaptive e-learning system.Finally, the functional modules and system architecture of the ontology-based adaptive e-learning system (OntoAES) were presented, the correlation and application pattern between various Ontology and system modules were studied. The adaptive e-learning steps and process of the OntoAES were discussed. Based on analysis of user learning behavior records, the analysis and definition of the potential learning resource relation pattern based on users’ use of log, the relation model and user preference model were acquired through information extraction and data mining technologies. Learning path information in Domain knowledge and user preference information could be refreshed; the model and method this paper presents were verified.The main contributions are as follows:1) Present the relation model of knowledge space, and method to build domain knowledge Ontology based on knowledge engineering;2) establish the Ontology of computer science domain and user model Ontology, present the architecture of the Ontology-based adaptive e-learning system;3) Define the adaptive e-learning rules based on Ontology, build the matching model between user preference ontology and study resource description ontology, study resource relation model and preference model based on data using.This paper provides the basis and guidance for establishing the ontology of the Education Semantic Web, and design and building of the adaptive e-learning system. The future work includes the perfection of the ontology and the function of the adaptive e-learning system, the study of the matching model in ontology as well as the rules of adaptive e-learning based on ontology, and to provide learners an intelligent and efficient learning system.更多还原