A Collaborative Deep Learning Model for Recommending e- Learning Resources for Intelligent Tutoring Systems

Big Data Analytics has been shown to be a powerful tool for curating personalized recommendation systems in diverse fields such as e-learning, movie and music etc. Due to the proliferation of e-learning applications and their associated advantages, it has become a difficult task to optimize e-learning resources according to user requirements, mostly due to the sheer volume of resources available. Individuals have varying preferences for learning resources, particularly depending on factors such as age, country, area, and specific needs. There is no definitive method to customize the recommendation system in order to assist customers in accessing the most optimal resources in the shortest possible timeframe. Collaborative learning has emerged as a successful method in machine learning approaches for developing recommendation systems. In this paper, collaborative filtering based deep learning (CFDL) approach with RMSProp has been developed for recommending e-learning resources. It has been shown that the proposed approach attains low Mean Absolute Error even a reasonably less iterations making it suitable for large dataset based recommendation systems.