基于行为分析的学习资源个性化推荐  

Personalized Recommendation of Learning Resources Based on Behavior Analysis

作　　者：聂黎生[1] NIE Li-sheng(School of Computer Science and Technology,Jiangsu Normal University,Xuzhou 221116,China)

机构地区：[1]江苏师范大学计算机科学与技术学院,江苏徐州221116

出　　处：《计算机技术与发展》2020年第7期34-37,41,共5页Computer Technology and Development

基　　金：国家自然科学基金(21776119);教育部产学合作协同育人项目(201902172045);江苏省社科基金项目(15TQB002)。

摘　　要：随着数字化学习资源规模急剧扩张,"知识过载"和"学习迷航"等问题限制了在线学习资源推荐的性能,学习者从海量的学习资源中选择合适资源的难度随之增大。针对传统推荐算法中存在的数据稀疏和学习资源个性化推荐精度不高等问题,提出了基于行为分析的学习资源个性化推荐算法。首先,构建学习者-学习资源评分矩阵;其次,挖掘学习者行为数据并将行为数据格式化融入到协同过滤个性化推荐过程;最后,计算学习者相似度并为待推荐学习者生成学习资源推荐列表。为验证模型的有效性,以"Live Course在线课程平台"数据为样本构建实验数据集,通过对比实验表明,该方法具有更高的推荐精度,能够更加精确和全面定位学习者的真实需求,实现学习资源个性化推荐。With the rapid expansion of the scale of digital learning resources,"knowledge overload" and "learning maze" and other issues limit the performance of online learning resources recommendation,and it is more difficult for learners to select appropriate resources from a large number of learning resources. Aiming at the problems of sparse data and low accuracy of personalized recommendation of learning resources in traditional recommendation algorithms,a personalized recommendation algorithm of learning resources based on behavior analysis is proposed. First of all,the rating matrix of learner learning resources is constructed. Secondly,the behavior data of learners is mined and the behavior data format is integrated into the collaborative filtering personalized recommendation process. Finally,the similarity of learners is calculated and the learning resources recommendation list is generated for the learners to be recommended. In order to verify the validity of the model,an experimental data set is constructed based on the "Live Course online course platform" data. The comparative experiment shows that the proposed method has higher recommendation accuracy,can more accurately and comprehensively locate the real needs of learners and achieve personalized recommendation of learning resources.

关 键 词：行为分析 学习资源 个性化推荐 协同过滤 推荐精度 

分 类 号：TP301[自动化与计算机技术—计算机系统结构] G434[自动化与计算机技术—计算机科学与技术]