The Prediction of Student Grades Using Collaborative Filtering in a Course Recommender System

Abstract

The increasing amount of data has made recommender systems necessary, especially with the internet and social media. Recommender systems refine the crucial information about users and items and process it in the most meaningful and helpful way. The course recommendation system offers students course suggestions, primarily non-technical and technical elective courses. In this study, student grades were predicted utilizing user-based collaborative filtering. How compulsory, technical elective and non-technical elective courses are handled in a system, the relationships of these courses were examined using up-to-date real-time data. The performance of the approach in grade prediction was evaluated according to the various performance metric groups. The outcomes show that preliminary analysis is critical when designing a course recommendation system. It is concluded that integrating a recommender algorithm into a student automation system can be innovative and yield more accurate guidance in students' success, e.g., elective course prediction performance with 96.4% precision is promising to guide students in finding the right area of expertise.