基于协同过滤的课程推荐模型*

通过对当前高校学生选课的盲目性的分析，提出了一种基于协同过滤的课程推荐方法。首先对课程进行聚类，构建无缺失的课程评价矩阵，在此基础上根据学生对相似课程的评分预测学生的兴趣爱好，为学生提供个性化的课程推荐。该方法在评分数据极端稀疏的情况下也可以为学生作出准确的课程推荐。最后通过实验验证了该推荐方法的实用性，可以有效地减少学生选课的盲目性。

Model of recommended courses based on collaborative filtering

By analyzing the blindness of students taking elective courses in the university, this paper proposed a recommended courses algorithm based on collaborative filtering. Firstly constructed a non-missing data curriculum evaluation matrix through the clustering of the courses. Then the method predicted the hobbies of students and provided personalized recommended courses for students according to the students rating on the similar courses. This method could make accurate recommended courses in sparse matrix for students.The experiment shows that this recommended method can well-targeted recommend courses for students and effectively reduces the blindness of taking courses for students.