【作者】 杨涛；

【导师】 索红光；

【作者基本信息】 中国石油大学 ， 计算机应用技术， 2007， 硕士

【摘要】 互联网的飞速发展改变着传统教育,同时也加快和推动了新型教育模式的出现。基于Web的个性化学习是在远程学习和个性化服务相结合的基础之上发展起来的。由于用户兴趣建模是个性化学习的基础和核心,所以本文围绕如何创建用户模型以及准确表示用户模型进行了详细研究,同时对用户兴趣模型如何在个性化学习系统中的应用也作出了相关研究。首先,分析了用户兴趣模型常用表示方法和兴趣建模的基本技术,并研究了用户长期兴趣和短期兴趣的变化和迁移,利用Web数据挖掘技术,针对用户的浏览内容和访问行为进行挖掘,自动地创建用户长期兴趣模型;另一方面,为了提高用户模型表示的精度,改进了基于关键词表示用户模型的常用方式,提出了一种基于概念表示用户特征项的方法,一定程度上消除了语义的多义性和歧义性。其次,利用已经建立的用户兴趣模型,调度适合用户个性的学习资源时,通过分析常用文档聚类方法以及研究互信息相关理论的基础之上,提出了一种基于互信息的Web文档聚类方法,对相关信息进行聚类,以达到更好地向用户推荐学习内容的目的。最后,对个性化学习系统中的主要功能模块进行设计和实现,并且通过实验验证了论文提出方法的可行性和有效性。更多还原

【Abstract】 The development of Internet changed traditional education, and accelerated the appearance of the new educational pattern at the same time. Personalized learning is based on the combination of distance learning and personalized service. Because the user interest model is the foundation and the core of the personalized learning, this thesis mainly discussed how to construct and express user’s interest model, meanwhile discussed the application how user’s interest models perform in the personalized learning system. Firstly, this thesis analyzed the common methods and the basic technology of user interest expressing, and discussed the change of user long-term interest and short-term interest. Then established user’s long-term interest model automatically through mining the pages and the behavior of user browsed based on Web mining. In order to enhance the precision of the user model expression, this thesis improved common mode based on the key word expression, and proposed a method based on the concept expression, which eliminated the semantic polysemy and the different meanings in the certain degree. Secondly, when dispatching the education resource based on established user interest model, proposed a method of web documents clustering based on mutual information through analyzing common methods of document clustering and researching mutual information theory. Finally, designed and explored the main function module of personalized learning system. And the experimental results indicate that this approach has better performance and validity.更多还原