Research on the Application of Item-based Collaborative Filtering Algorithms in MOOC

Abstract

In the learning process, the recommendation system can recommend learning resources that are in line with the learning situation of the learning objects and help the learning objects to learn more easily and naturally. By analyzing the characteristics of item-based collaborative filtering algorithm, this paper applies it to the recommendation system of mooc learning resources to avoid the possible defects of this algorithm, such as low performance and offline processing in other scenarios, and gives full play to its good real-time performance and high recommendation efficiency. By analyzing the log files of imooc, this paper obtains the learned learning resources of the learning objects in imooc, takes the first 80% data of historical records, obtains the recommended list of learning resources through recommendation algorithm, and then compares the data of the last 20% of historical records. Finally, the algorithm is optimized according to the accuracy of the recommendation algorithm in the experiment and the actual learning scene.