面向MOOC平台的课程推荐研究综述

Course recommendation for MOOC platform: A review

随着互联网技术和在线教育的飞速发展,目前我国已出现大量线上教育平台,但这些在线教育平台相互间信息不能共享,导致课程信息冗余过载,用户选择困难.本文综述了近年来课程推荐方面的研究进展,首先介绍了课程推荐中的相关概念并给出了系统框架;然后围绕课程建模、用户建模、核心算法3个方面进行探讨,重点综述了5类算法:内容推荐、协同过滤、混合推荐、深度学习推荐和多模态融合推荐,并分析了数据集、实验方法和评价指标;最后对个性化课程推荐技术进行了总结和展望.

With the rapid development of internet technology and online learning, a number of online education platforms have emerged across China. However, these online education platforms cannot share information with each other, which leads to redundancy and overload of curriculum information and difficulties for users to choose. This article reviews the research progress of course recommendation in recent years. First, the related concepts in course recommendation are introduced and the system framework is given. Then the three aspects of course modelling is discussed, including user modelling, and the core algorithm of the recommendation system,etc. Five types of algorithm are presented: the first three categories are content recommendation, collaborative filtering and hybrid recommendation;while the rest two are deep learning recommendation and multi-modal fusion recommendation, and the data set. In the end, the experimental methods and evaluation indicators are summarized, and the personalized course recommendation technology has also been projected.