基于半自动编码器的学习资源推荐

Research on learning resource recommendation based on semi-autoencoder

为提升教学资源库的个性化学习资源推荐效果,以软件技术国家专业教学资源库中的学习数据为基础,结合学习行为特征,构建高等教育个性化学习体系框架.基于半自动编码器的学习资源推荐模型,运用学生和学习资源等辅助信息,缓解学习资源推荐的数据稀疏问题,根据学生学习偏好分析结果为学生推荐合适的学习资源.实验结果表明,该学习资源推荐模型优于传统推荐算法,能有效提升资源推荐准确度.

In order to improve the recommendation effect of personalized learning resources in the teaching resource database,based on the learning data in the national teaching resource database of software technology majors and combining with the characteristics of learning behaviors,the framework of personalized learning system in higher education is constructed.The learning resource recommendation model based on semi-automatic encoder uses auxiliary information such as students and learning resources to alleviate the data sparse problem of learning resource recommendation,and recommends appropriate learning resources for students according to the analysis of students’learning preferences.Experimental results show that this recommendation model is superior to other traditional recommendation algorithms and can effectively improve the accuracy of recommendation.