In spite of an increasing development of virtual and distant applications which use the advantages of multimedia and the Internet for distance education, learning by means of such tutorial tools would be more effective if they were specifically adapted to each user needs. This paper proposes a knowledge representation model which judiciously serves the remediation process of an intelligent learning environment in interaction with students during virtual learning activities. By means of experimental results obtained thanks to practical tests, we show that our knowledge representation model facilitates the planning of a tailored sequence of feedbacks that significantly help the learner.