Vyrodov O. Hypermedia and neuronet methods and models for computer-based didactic systems

The dissertation is dedicated to problems of development of mathematical support for computer-based didactic systems. The approaches and mathematical models used in modern tutoring systems have been analyzed. A model of computer-based didactic system that is invariant relative to the domain has been suggested. A method of control of learning information presentation has been developed. The method adapts to the student's knowledge level and interests. The application of fuzzy logic to student modeling has been improved. A student model based on neural network has been developed. The student model requires no prior student's knowledge testing. The task of student's domain browsing pattern recognition has been defined more precisely and solved. The method of determining of students domain learning strategy has been developed. The task of automation of computer-based didactic system development has been investigated. The procedure and structure models of software tool for automation of computer-based didactic system development have been derived.