Abstract
One of the biggest challenges in education is teaching mathematics, especially to children. It has been proven that difficulties students face when learning basic mathematics are often the result of previously acquired misconceptions. These misconceptions prevent the student from understanding new concepts and will eventually create a psychological barrier that prevents the student from learning more advanced mathematics. The conventional classroom environment does not provide the teacher with the most efficient means to detect and correct such misconceptions. The goal of our research is to develop an e-learning system for basic mathematics that is capable of providing each student with personalized content to overcome these misconceptions. The system uses a multi-agent architecture to monitor the activity of the student while simultaneously observing and modeling the student's knowledge and misconceptions. Lessons and exam questions are chosen dynamically by the multi-agent system to cover the prerequisites of new lessons depending on the profile of the user.