基于学生特征模型的教育云资源推送技术

发布时间：2018-09-08 11:24

【摘要】：近年来,由于互联网不断地深入人类社会的各个领域,网络数据越发泛滥,从而云计算、分布式存储等大数据处理技术快速地发展。许多云系统都在积极筹备、建设之中,面对日益庞大、复杂的资源池、个性化推送技术的应用已经变得势在必行。教育领域也不例外。随着教育信息化程度的提高,人们已逐渐不再仅仅去图书馆、书店寻找自己需要的教育资料,而是更多地通过互联网络查询、检索想要的数字资源。不仅如此,在线教育、公开课堂等网络教育也越来越流行。而在海量的教育资源中,如何快速有效地满足用户的个性需求,成了“教育云”系统服务的重要内容。基于现状,本文提出了一套面向学生的个性化推送方案。以满足学生对教育资源的需求。本文首先对分布式存储系统和几种主流的个性化推送技术进行了研究、分析,比较它们的优缺点。接着,分析教育云系统的用户群和资源池的特征,结合基于内容的推送技术的特点,建立学生用户与教育资源的数学模型。学生模型以学生的知识广度、年级、成绩等特征作为变量；资源模型则以教育资源的知识面、难易度等特征为变量。在建立的模型的基础上,提出了一套教育云系统的个性化推送方案。整个推送方案包含了特征的提取、知识广度的匹配和知识深度的匹配等主要模块。并逐一对各模块进行了设计、实现。最后,在以Hadoop集群为主体的分布式存储平台上,初步实现了这个教育云系统的个性化推送方案。论文提出的个性化推送方案可以作为教育云系统个性化服务的组成部分,应用于以Hadoop集群为存储系统的教育平台。
[Abstract]:In recent years, because the Internet has continuously penetrated into various fields of human society, the network data has become more and more widespread, thus big data processing technology such as cloud computing, distributed storage and so on has developed rapidly. Many cloud systems are actively preparing and building. In the face of the increasingly large and complex resource pool, the application of personalized push technology has become imperative. The field of education is no exception. With the improvement of educational informatization, people are no longer just going to the library, bookstores to find their own educational materials, but more through the Internet query, to retrieve the desired digital resources. Not only that, online education, open classroom and other network education is becoming more and more popular. In the mass of educational resources, how to meet the needs of users quickly and effectively has become an important part of educational cloud system service. Based on the present situation, this paper puts forward a set of personalized push scheme for students. To meet the needs of students for educational resources. In this paper, the distributed storage system and several popular personalized push technologies are studied, and their advantages and disadvantages are compared. Then, the characteristics of user group and resource pool of educational cloud system are analyzed, and the mathematical model of student users and educational resources is established by combining the characteristics of content-based push technology. The student model takes the characteristics of students' knowledge span, grade and achievement as variables, while the resource model takes the characteristics of educational resources, such as knowledge, difficulty and so on, as variables. Based on the established model, a personalized push scheme of educational cloud system is proposed. The whole push scheme includes feature extraction, knowledge breadth matching and knowledge depth matching. Each module is designed and realized one by one. Finally, on the distributed storage platform with Hadoop cluster as the main body, the individualized push scheme of the educational cloud system is implemented preliminarily. The personalized push scheme proposed in this paper can be used as a part of the personalized service of the educational cloud system and can be applied to the education platform with Hadoop cluster as the storage system.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP391.3;TP333

【参考文献】