Designing, implementing and validating a recommender system represents a challenge that has been tackled within many e- Learning platforms. Still, each proposed approach has to take into consideration the particular underlying data workflow and wrap-up together appropriate mechanisms (i.e., memory based, model-based or hybrid) to obtain an effective recommender system. This paper presents a custom-designed validation procedure for a recommender system that has been previously developed and integrated into Tesys e-Learning platform. The recommender system is evaluated in terms of correctly recommending concepts of study, in accordance with the learner's particular knowledge level measured by previously taken tests and the knowledge of all other learners that have answered quizzes in the same learning context. Experimental results show an increase of the percent of correctly recommended concepts over time. Within the validation mechanism, a comparative analysis with two test scenarios shows that proposed recommender has an increase in value for the percentage of correctly recommended concepts. The custom proposed evaluation procedure may quickly evaluate further improvements of the recommender system.