协同过滤算法在高校课程推荐中的应用研究

Researches on Collaborative Filtering in University Course Recommendation

伴随高校信息服务系统的深入应用,大学生对各种选修课程的选择呈现出更具个性化和社会化的趋势,课程推荐服务成为学生在使用选课系统和慕课等平台的过程中所依赖的重要服务.传统推荐算法主要应用于针对电子商务领域的商品推荐服务,而课程推荐则主要部署于服务学生的各种课程服务系统中,二者所依赖的背景知识和目标存在较大区别,因此本文针对课程推荐服务中的知识特征,通过对比分析协同过滤算法主要工作过程在高校课程推荐与商品推荐中的差异,提出了对传统协同过滤算法的改进和完善措施,以提高课程推荐的整体推荐质量.

With the further application of university information service systems, the student's way of selecting optional courses shows a trend of more personalization and socialization, and the course recommendation is becoming a dependent service for students during interacting with course selection systems and MOOC platform. The classical recommendation approaches have been mostly applied for product recommendation in the field of e-commerce, while the course recommendation is widely employed in course service systems for students, there are more differences about the background knowledge and objective involved in both types of recommendation, therefore according to the knowledge feature of course recommendation, by comparing the differences existing in the key steps of collaborative filtering applied in university course recommendation and product recommendation, some improvements aiming at collaborative filtering are proposed to enhance the overall service quality of course recommendation.