Learning style is considered as preferred manner in which learners acquire effectively new knowledge. Individual's learning style contains mainly his preferences, his expectations and his needs. In addition, it represents a key asset in improving teaching methods that meet learners' needs. During the last decades, the identification of learning style methods has been relied on two basic measure methods: collaborative and automated ones. In this paper, we propose automatic test PRFSLM is based on Felder-Silverman learning style model (FSLM) that helps instructors to create and personalize their placement task's items. Our proposed ontology-based automatic method aims (1) to assess the learner's prerequisites and also (2) to provide explicitly an extensive knowledge about learner preferences that could predict his own learning style at the beginning of an online training.