Adaptive Predictive and Recommendation System Based on Learners Style

Abstract

E-Learning development has been a widely adopted methodology nowadays. But there are main risks for learners that they cannot understand in a single stretch of learning. The learning style of every learner varies according to the knowledge and learning object level. Several points have emerged regarding the conception, development and maintenance of E-learning solutions. In our proposed system, Felder Silverman learning style modules had been used to analyze the learner style and the predictive mechanism is enhanced using machine learning methods. The risk of learner will be reduced with the implemented tool. Deep learning can predict the learner style trained labeled data inputs and the tested outputs produce a visualization model of the learner score. The learner scores and the recommendation strategy are added to make learning easier using E-learning technology. Thus the learners can be known as (visual, aural, physical, logical, etc.). This reduces the risk of online learners and the learning objective is well known by the user itself using the implemented methodology. System act as an instructor with the proposed system and an artificial intelligence system is added for a systembased instructor. The statistical test showed that both the behavioral intervention and recommending intervention is based on earlier prediction played a positive role and learning engagement.