AI-Driven Content Recommendation in Learning Management Systems (LMS): A Hybrid Filtering Approach

Abstract

The incorporation of technology in teaching has drastically changed how learners engage with scholarly resources for
a particular subject. Today's Learning Management Systems contain innumerable resources that require intelligent
recommendation systems to optimize learner interaction and achievement. This study focuses on creating an AI-based
content recommendation system using hybrid filtering, where the recommendation is based on both content and user
activity within the LMS. The main focus of the system is to mitigate the cold-start problem and over-specialization of
traditional recommender systems. A prototype was built with a custom LMS that incorporated a hybrid recommend
system which was tested in an experimental setting with students from a local university against traditional filtering
techniques. Data regarding precision, recall, user engagement, and satisfaction was collected and analyzed. Results
from the study showed the integrated use of diverse AI techniques more accurately met the user's needs regarding
educational content access as learners rated their experience more positively. The research demonstrates the significant
advantages that fusion models present in the modern educational context while outlining the direction of research that
adaptative learning systems require.