Today various e-learning systems which spectrum and number is growing from year to year are widely used in the learning process. However, their competitiveness and use in practice depends not so much on system functional capabilities, as it was at the initial stage of integration of such systems into the learning process, but on system flexibility in terms of its suitability to characteristics and needs of particular learners. In order to achieve the mentioned flexibility, as a rule, the student model is used. It stores information concerning progress of the learning process, state of knowledge and skills of the learner, his/her cognitive, emotional and psychological characteristics. The paper is devoted to the student model which in the autumn of 2010 was implemented in the intelligent knowledge assessment system allowing the evaluation of learner's knowledge structure on the basis of concept map based tasks. The paper describes the application of the model to the selection of such degree of task difficulty which is the most suitable for learner's knowledge level and study style, as well as considers the technique for the experimental evaluation of adaptation mechanism which uses data about a student.