Posadska I. Methods and models of content and control tests structuring in automated learning systems

The scientific and applied task of the dissertation work is the development of new methods of the learning course domain structuring including its in the learning system and the development of technologies for knowledge automated control using its formalized representations. The object of research is the process of automated learning. The subject of the research is models and methods of content and control tests structuring in automated learning systems. The aim of the research is to improve the quality of automated learning using new methods of ALS content structuring that will provide the possibility to apply various forms and methods of knowledge automated control. For the first time it was proposed in the automated learning systems to structure the learning courses in the form of a domain model represented by UML diagrams and formal logical theories. The work technology with the automated learning system, which supports not only the functions of a student and a tutor, but also a domain expert and a knowledge engineer, has been further developed. For learning systems, the formalization of the learning courses content in the form of UML diagrams was further developed by introducing strategies sets for learning course domain constructing. For the first time, methods of domain fragments integration, described in UML diagrams, into a conflict-free diagram of the learning course are proposed. The process of selecting of optimized variants of learning course domain has been further developed, based on the application of expert assessments of the selected indicators of the domain representation quality and the method of analyzing hierarchies. Based on the research, a subsystem of knowledge automatic control was developed and implemented for measuring of the students' activity using the learning system resources, which allowed for rapid assessment of the student's knowledge level.