Inteligent tutoring systems are e-learning systems, developed with the goal of emulating the tutoring teaching model, a highly efficient learning environment based on complete adaptation of the learning process to the needs of one student. The goal of such systems is that the computer system behaves in an intelligent way, e.g. that it adapts to the current needs and knowledge level of the student. This is achieved in various ways such as the application of artificial intelligence or data mining methods. Intelligent tutoring system are developed mostly for teaching in well defined knowledge domains such as mathematics, physics, etc., but there are many other not so well defined areas that also need to be taught. These areas are called ill-defined domains, where the teacher (expert) prepares the learning process based on his/hers knowledge and experience as well as the current „intelligence“ implemented in the system. By applying data mining methods to data about students' interactions with the system we can discover useful information about their learning processes and combine this information with the current system functionality in order to improve it's efficiency. This thesis proposes a new model of an existing intelligent tutoring system which, through the implementation of aforementioned methods, offers the student suggestions on which steps to learn next in order to streamline his/hers path through the knowledge domain. The key components of the presented system are: a communication layer for communication with data mining tools, a clustering model discovery and selection module and a high-utility frequent sequential patterns discovery module. The new tutoring model of the intelligent tutoring system then offers suggestions to the user in the form of hyperlinks to other knowledge units that are best learned before or after the unit the student has selected. The described system is implemented as a web application called DITUS (Department of Informatics TUtoring System). The validitiy of the proposed model was verified through an experiment wich determined its viability and efficiency.