Using Data Mining for Learning Path Recommendation and Visualization in an Intelligent Tutoring System

Since the creation of Web-based learning systems researchers have tried to make them adaptive to various characteristics of learners in order to increase the quality of learning. The application of data mining (DM) methods on learning system logs is often used as a basis of intelligent tutoring systems (ITS) that have the ability for automatic adaptation of some aspect of the learning process. Such a system was developed at our institution in the previous years and has been used in a number of courses. To improve our system we added an integration application to create a continuous feedback loop with a DM tool. Our goals, from the student's perspective, are to improve the quality of knowledge acquired by students as well as shorten the learning time by offering recommendations based on mined patterns of learning paths through the knowledge domain. From the perspective of teachers, our goal is to create a rich data visualization system (in the form of a Web application using standard Web technologies to create visualizations) to give them new insight into students' behaviours. In this paper we present the structure of our system and the research design that will be used to verify its results.