Integration of Learning Analytics into Learner Management System using Machine Learning

Abstract

The demand of e-learning is constantly increasing at a rapid rate for the educational institutions. Web-based Learning Management System (LMS) is one of the main components for the e-learning system. There are multiple benefits of using LMS including cost reduction, content management, flexibility, and many more. Despite of theses significant benefits of using LMS, traditional LMS system cannot supports with modern learning needs. In particular, extracting useful information from the huge educational data and analysis and interpretation that information is challenging. Learning analytics can effectively support these needs in terms of predict learner performance, engagement and potential problems. This research presents learning analytics using machine learning techniques and considers its integration into LMS. The approach considers various indicators such as assessment score, gender and age for the prediction. This certainly supports organization to undertake actionable decisions as preventive measures for the overall teaching and learning support. We have considered a widely used learner data sets to demonstrate the applicability of our approach. The result shows that Decision Tree shows the highest accuracy among the chosen three ML algorithms. We have observed that average grade for a given course acts as an important indicator to predict over all outcome.