基于深度学习的在线教学推荐系统设计与研究

Design and Research of Online Teaching Recommendation System Based on Deep Learning

智能移动终端与无线网络技术日益普及，使得人们的学习与生活不断变化，传统的在固定时间、固定地点的学习方式已经不能满足人们及时、高效的学习需求．将教学资源整合到移动平台，并利用碎片化、微型化、多任务和浅层识别等特征的学习新方式已成为当前研究的热点．但现有的大部分系统均是将教学资源从原有的网络化平台移植到移动端，简单的进行了分类归类，并没有进一步挖掘教学资源所对应知识点和学科群之间的关联关系．因此，从实现个性化学习的实际问题出发，基于深度学习理论，利用卷积神经网络实现教学资源的特征表达、提取与聚类．将教学资源对应知识点的特征属性进行反复抽象与迭代，实现精准高效的移动平台教学资源推荐系统．

The growing popularity of intelligent mobile terminals and wireless network technology has made people＂ s learning and life constantly changing. The traditional way of learning at a fixed time, fixed location has been unable to meet people＇s timely and efficient learning needs. The new learning way to integrate teaching resources into the mobile platform and utilize the features of fragmentation, miniaturization, multi - task and shallow identification has become a hot topic in current research. But most of the existing system is to transplant the teaching resoures from the original newtwork platform to the mobile terminal, simplely classify, not furtherly excavate the relationship between the knowledge points of teaching resoures and its related subject group. Therefore, from the practical problems of personalized learning, based on the theory of deep learning, we use convolution neural network to realize the feature representation, extraction and clustering of teaching resources. The characteristics of knowlege points corresponding to the teaching resources are repeatedly abstracted and interated to achieve a precise and efficient teaching resource recommendation system for mobile platforms.