AbstractConsidering the increasing importance of adaptive approaches in CALL systems, this study implemented a machine learning based student modeling middleware with Bayesian networks. The profiling approach of the student modeling system is based on Felder and Silverman's Learning Styles Model and Felder and Soloman's Index of Learning Styles Questionnaire. The questionnaire was adapted to Turkish for this experimental study conducted with respect to the visual/verbal and active/reflective dimensions of the model. A topic in EFL was chosen for the learning content design, which was also carried into the digital domain and remastered as separate learning scenes for different learning styles. Computer software was also implemented to carry out the experimental learning processes. A quasi-experimental pre-test, post-test design was conducted with 46 volunteers, with 23 students assigned each to a control and an experimental group to compare academic achievement between student-based learning and conventional computer-based learning. No significant difference was found in academic achievement between the control and experimental groups after the experimental treatment. The diagnostic performance of the proposed student modeling system was also compared with performances from similar studies. This student modeling system had a successful prediction rate of 41% on the visual/verbal dimension and 54% on the active/reflective dimension, respectively.Key WordsAcademic Achievement, Bayesian Networks, Computer Aided Language Learning, Felder and Silverman's Learning Styles Model, Student Modeling.The development of advanced communication techniques and widespread use of the Internet makes knowledge acquisition without geographic limitations much easier. These changes have triggered the transformation of student needs, following the path from traditional synchronous education to semi-synchronous and blended e-learning environments that are mostly realized by the clicks of a common mouse on a common software browser. While the teacher's role shifts from solely unidirectional teaching to facilitating the use of learning materials, the student's role also shifts from passive learning to taking responsibility for the learning process and continuous selfassessment. As a result, during e-learning, the student usually becomes the decision maker. There are, of course, pros and cons of having such individual responsibility. The positive and negative effects of e-learning on online students have been studied in detail by Cantoni, Cellario, and Porta (2004). One of the most important negative effects of e-learning is the student's solitude during learning sessions because of limited interaction. In such cases, student dropouts become a common reason for failure when the learning environment does not meet their expectations (Katz, 2002; Levy, 2007), while student-centered learning environments can enhance students' motivation by using intelligent computer systems.Even when students and teachers synchronously collaborate for a common purpose, they usually exist in different psychological, environmental, and geographical conditions. When the student's self-control is in question, it is crucial that the interaction between the student and the learning system must be maintained at a high level. The variety of quite different, individual methods used by students during their learning must not be overlooked when interaction is what matters (Baldwin & Sabry, 2003).It would not be surprising to think that the underlying component of an e-learning system is more or less sophisticated computer software designed to work with specific hardware requirements. It would then not be surprising that the processing superiority of computers over human beings lies at the heart of student-centered learning in the form of customized service possibilities. There are many customization and personalization systems installed in end-user computer software solutions nowadays. …