Considering recent explosion in data and information amount, applying faster and more accurate solutions for information mining and extraction for different users is inevitable. One of the newest issues which is designed and developed based on web technologies is ELearning. The main goal of this area is to provide learning services for users in any place anytime. It is obvious to except that the learning styles of the different users are varied. The point mentioned in the literature is that the training quality is related to the users' learning styles. Besides the questionnaires which are designed in order to recognize this, other approaches for this aim are also proposed by researchers in computer science. One of the proposed solutions to improve the learning courses is to justify; MARGIN: 0cm 0cm 0pt; unicode-bidi: embed; DIRECTION: ltr" Keywords: Learning Style, MFMC Algorithm , Mutual Information , Markov Model, Naïve Bayes Algorithm