基于GRU和课程关联关系的推荐模型
2020-08-13卜祥鹏



摘 要: 如今以MOOC平台为主的在线学习网站课程资源丰富,数据量庞大,存在明显的信息过载问题。用户如何在海量数据中快速找到合适的课程是一个难题,而个性化的课程推荐是解决这一问题的有效方式。考虑到用户学习的课程存在先后顺序,具有明显的时序性,所有我们使用门控循环单元(GRU)从用户学习的课程序列中挖掘其存在的规律和特点。另外,为使推荐的结果更加准确,我们考虑结合课程之间的关联关系,提出了基于GRU与课程关联关系的推荐模型。最后,我们在真实的MOOC数据集上进行大量实验,实验结果表明,与传统的协同过滤算法以及仅考虑课程关联关系的算法相比,我们提出的算法性能更好,推荐的准确度更高。
关键词: 课程推荐;门控循环单元(GRU);课程关联关系;MOOC
中图分类号: TP183 文献标识码: A DOI:10.3969/j.issn.1003-6970.2020.06.029
本文著录格式:卜祥鹏. 基于GRU和课程关联关系的推荐模型[J]. 软件,2020,41(06):137142
【Abstract】: Nowadays, the online learning websites mainly based on MOOC platform are rich in course resources and large in data volume, so there are obvious problems of information overload. How to find the right course quickly in the mass data is a difficult problem. One of the most effective ways to solve this problem is to use personalized course recommendation. Considering that the courses learned by users are in order and have obvious timing, we use Gated recurrent unit (GRU) to mine the existing rules and characteristics of the courses learned by users. In addition, in order to make the recommended results more accurate, we consider the correlation between courses and then propose a recommendation model based on the GRU and course association relationship. Finally, we conduct a lot of experiments on real MOOC dataset, experimental results show that the proposed algorithm has better performance and higher recommendation accuracy than the traditional collaborative filtering algorithm and the algorithm that only consider the course association relationship.
【Key words】: Course recommendation; Gate Recurrent Unit(GRU); Course association relationship; MOOC
0 引言
隨着科学技术和在线服务业的发展,出现了以慕课(MOOC)平台为代表的在线学习网站。与传统的教学方式相比,在线网站学习具有课程资源丰富和灵活度大等优点,能够满足各种不同类型的用户对学习资源的需求,得到了人们的广泛认可。然而随着在线学习网站的普及,信息过载的问题日益突出,海量的学习资源使用户感到困扰,如何快速找到适合自己的学习资源成为一个巨大的挑战。为解决这一问题,人们提出了个性化课程推荐方法,即根据每个用户自身的特点,为其推荐适合的课程,这也成为当前教育数据挖掘领域主要的研究问题之一。……
