Integrating Machine Learning into Learner Profiling for Adaptive and Gamified Learning System

Abstract

The success of an adaptive learning system depends on the learning content. Each student seeks an environment that is suitable for his needs, with personalized and adaptable content that allows him to have a more successful and meaningful learning experience. Learner profile is a structure comprising direct and indirect information of learner’s background, objectives, interest and preferences. Taking a leaner’s profile into account while designing courses is beneficial, and profile modeling is an essential method that seeks to provide a comprehensive representation of all factors linked to the user's attributes. In this paper we propose a machine learning model for predicting learner profile. It serves as a basis to a suitable user-centered adaptation of gamification and content. The potential of our model is considering both the player and learner contexts by integrating learners’ interactions, preferences, troubles and cognitive capacities. We tested the efficiency of our contribution in a gamified learning environment called “Class Quiz”. We used a dataset of 1000 examples to develop classification models by combining several techniques.