【作者】 南菊红；

【导师】 余雪丽；

【作者基本信息】 太原理工大学 ， 计算机应用， 2003， 硕士

【摘要】 本文根据中国现代远程教育资源建设的现状，管理模式的不足，研究如何高质量地建设分布式的资源环境，以适应现代远程教育的发展，避免教育资源的交叉重复建设和浪费。现代远程教育资源建设存在信息壁垒问题，数据缺乏合理的架构，这是造成资源重复建设和浪费的很重要原因。而资源的建设与共享是一项系统工程，系统要正常运行，必须遵循规范化的合理的设计模式和体系结构。所以对于远程教育系统资源建设来讲，如果开始就能够自上而下地遵循规范化的途径有序地进行，就可以提高数据管理与共享的效率。随着网络技术的发展以及教育业务的拓展，分布式数据库技术的发展，无疑为教育资源的建设提供了技术支持。所以，采用分布式数据库技术构建远程教育资源框架DLRF(Distance Learning Resource Framework)是必要和可行的。为提高系统的可靠性，可用性，系统的本地自治性，并减少网络流量，以分层分段作为数据模式设计的指导思想，这是解决复杂问题的常用方法，分布式资源体系架构也不例外。层次分明、条理清晰的架构可以更好地管理分布式系统，有效地协调分布在各处的节点，形成完整的分布式统一体，最终达到整个系统的良好运行，这是DLRF的关键目标。 在资源环境设计时，采用自顶向下的方法，从需求入手，对业务需求，功能需求，非功能需求，外部约束等几个层次进行分析，为具体实现方案的提出准备条件。根据需求分析结果，设计出分布式资源管理环境实现方案，并进行部分验证。由于教育系统的多元化，使得不同的系统对教育词汇的定义往往产生分歧，针对这种现状，我们本着先实现共知，后实现共享的原则，运用本体描述领域中的关键术语，建立领域词汇表，为信息共享创造条件。对以后信息深层提取也有好处。选择本体描述也是从经济方面考虑，标准的方法总是比非标准方法要合理、经济。并且给出一种新颖的信息互操作和共享层次模型，这种模型虽然是为消口太原理工大学硕士研究生毕业论文除远程教育系统中的语义冲突而提出来的，但它同样适合于其它应用领域，如数据挖掘、数据仓储、电子图书馆等。考虑到系统要求具有较高的扩展性，可用性以及数据处理能力等因素，对教育资源使用机群技术来构建系统平台。系统从逻辑上分为7个分区，每个逻辑分区完成不同的功能。数据管理主要采用分布式管理与集中式管理相结合的模式。采用分层和分段两种方案来实现数据管理。数据层的组织方式为XML和oracle关系数据库相结合。 在分布式资源环境中，离开元数据的资源将是一盘散沙，将无法提供有效的检索和处理。所以，引入元数据管理资源信息是一种有效的方法。元数据帮助检索和确认所需要的资源，支持资源的存储和使用管理，数据元素除比较全面的资源描述信息外，还包括权力管理、资源评鉴、使用管理等方面的信息。其中部分元数据描述以XML为媒介。目录管理采用分布式目录方案，即本地目录和集中式目录相结合的方法。提高了系统的执行效率和可用性，降低了更新目录的复杂性。为减少不相关数据的访问，减少网络流量，对需求得到的全局概念模型根据应用特性，运用谓词方法、BEA算法对其进行分割，得到本地概念模型。然后在深入研究了现有的分布算法后，对分布算法作了可行的适应性修改，在事务特性模型的基础上，依据事务模型把分段分布到节点上。 最后，对本课题进行总结，提出了需要继续深入的工作。更多还原

【Abstract】 The thesis has analyzed the status and shortage of the resource environment in the distance learning. In order to effectively avoid the rebuilding and waste of the distance learning resources and to promote resource management, interoperability and extensibility capabilities, as well as to allow them to be combined for performing more complex operations, a standard reference architecture is needed. We have presented to apply the distributed database technology into the DLRF (Distance Learning Resource Framework) system. The thesis has mainly studied how to design and build a well-structured distributed resource framework. Now there are information gulfs in the distance learning resource systems. The degree of information sharing among different systems is low and the information lacks of proper structure. It is one of the reasons that have caused the rebuilding and waste of resources. But the resource construction and information sharing is a systemic engineering. In order that the system will run well, the process of system design must abide by formal design schema and architecture. Thus, if we can comply with the norm during the design process of the distance learning resource environment, it can greatly improve the information management and sharing capability. With the development of network technologies and distributed database technologies, they have provided the technology support for distance learning resource construction. Consequently, distributed database architecture is a natural choice for it. It is necessary and feasible to applied the distributed database technologies into our DLRF system considering the increasing expansion of education business and increasing resources. During the data model design process of the DLRF, we abide by the hierarchy and fragmentation principle in order to improve the reliability, availability, and local system autonomy of the system and to reduce the network flow. The hierarchy and fragmentation principle is a good means for solving complicated problems such as DLRF. The clear hierarchy framework is better for managing distributed system, effectively coordinating the sites residing at differentlocations, and then forming an integrated and well running distributed system. It is the key purpose of the DLRF.During the design process, we have firstly analyzed the requirements of the system, including business analysis, function analysis, non-functional analysis and the external constraints. Then we have expounded the design scheme involving the architecture and components and hierarchy information management method. In view of the diversity and multi-model of education system, it is often that different systems with different purposes may reasonably come to different interpretations about the same terms. Abide by the principles of "to first realize that different ones have the same cognizance about the same terms and then to realize information sharing", we use ontology to describe the critical terms in distance learning and to structuralize and to provide equivalence criteria, and at the same time it can effectively reduce the semantic conflicts and facilitate the interoperability among different data sources. It also has created conditions for in-depth information abstraction. The motivation behind the ontology practice is economic: there is benefit to doing things in a standard way and cost to doing things differently. Then we also have put forth an interoperable and sharing hierarchy model for distance learning. Though this model is put forward for distance learning, it also can be applied to other application domains, such as Web-based databases, digital library, data warehouses, data mining, etc. Considering the system performance demand, we have chosen a group of servers or cluster to establish the resource platform. The system are classified into 7 logical categories, each one executes different function. The data management has adopted the means of integrating the distributed control mechanism with the central control mechanism. We have applied the h更多还原