齐　婷,佟国香,2

(1.上海理工大学 光电信息与计算机工程学院,上海　200093;2.上海市现代光学系统重点实验室,上海　200093)

基于改进的混合模式个性化选课推荐技术研究

齐婷1,佟国香1,2

(1.上海理工大学 光电信息与计算机工程学院,上海200093;2.上海市现代光学系统重点实验室,上海200093)

摘要针对高等学校学生选课系统中存在的缺乏个性化课程推荐、选课效率较低的问题,通过对个性化推荐技术的分析研究,提出了基于内容、项目及用户属性的改进混合模式算法,并将该算法应用到选课系统中,用MACE数据集对算法进行验证。结果表明,该算法解决了个性化推荐技术中的冷启动问题,相关指标有明显提高,实现了课程与新课程的个性化推荐,并减少了选课的盲目性。

关键词个性化推荐;混合模式;相似度;用户聚类

Research on Improved Personalized Courses Recommendation Technology Based on Mixed Mode

QI Ting1,TONG Guoxiang1,2

(1.School of Optical-Electrical and Computer Engineering,University of Shanghai for Science and Technology,

Shanghai 200093,China;2.Shanghai Key Laboratory of Modern Optical Systems,Shanghai 200093,China)

AbstractProblems of lacking in individualized curriculum recommendations and inefficiency exist in current course selection systems of institutions of higher education.In allusion to these limitations,this paper presents a improved mixed model algorithm based on the content,project and user attribute-value through analysis and study of personalized recommendation technology.The proposed algorithm has been successfully applied to the elective system.Experimental results indicate that the proposed approach can solve cold-start technology in personalized recommendation algorithm,improve the related indicators significantly,achieve a personalized recommendation and new courses recommendation and reduce the blindness by the MACE data sets.

Keywordspersonalized recommendation;mixed mode;similarity;user clustering

随着学生在教学过程中主体地位的突出,学生选课已成为学生个体化发展的重要途径,而多数高校在选课制度实施过程中普遍存在课程结构设置不合理、选课方式不完善、选课指导体系不健全等问题[1]。学生不能结合自身的专业和兴趣进行选课,选课缺乏目的性和针对性;选课制度不利于学生的个性发展,也不能为学生以后的工作带来良好的指导,从而出现了专业与职位不对口的现象[2]。有鉴于此,本文将个性化推荐技术应用于选课系统中,根据学生自身的状况、学习需求、兴趣偏好以及职业规划等,为学生提供个性化课程推荐平台,从而避免学生选课的盲目性和跟风现象,提高了课程资源的利用率和选课质量。……