During the research, more than 100 sources were processed: among them books, textbooks, scientific articles, thematic and project sites, discussions, standards, patents. Source codes of open source projects related to the studied subject area in such programming languages as Java, C/C++, R, Python, Julia, Go, Prolog, Common Lisp were considered. Through the development of prototypes, preparation and analysis of data, implementation of test and training tasks, the approbation of models, packages, frameworks and libraries, comparison of their properties and limits of application was performed. The results and artifacts of testing the researched tools and methods can be found in the open repository of the master's thesis [20].
As a result of the study, the ways of intellectualization of existing EL systems through the application of ontological KB and AOD were determined. Integration of EL with such subsystems as ITS and CBE.
ITS supports a detailed model of a student's skills and learning history, which is represented by an ontology or knowledge graph. This model helps the system track the current level of knowledge and learning trajectory. ITS also includes ontologies or knowledge graphs specific to the discipline being taught. Such representations capture the structure of the domain and the relationships between concepts, allowing the system to provide contextually relevant recommendations. ITS can tailor content and learning strategies to meet individual student needs and progress. She can adjust the difficulty of the questions, provide additional explanations and offer additional materials as needed. ITS provides instant and constructive feedback to students. Feedback can include explanations of correct and incorrect answers, hints on how to solve the problem, and suggestions for improvement. It provides an opportunity for self-evaluation, student progress is tracked to identify areas that need improvement.
CBE provides competency maps or knowledge graphs that provide a representation of the relationships and dependencies between competencies. It supports individual learning paths for each student based on their current competencies and desired learning outcomes. This includes recommendations on the sequence of competencies to be mastered and the learning resources required to be used. CBE can automate competency assessments, both initial and final, to determine a student's level of mastery of specific competencies. These grades can be adapted to the student's demonstrated knowledge and skills. CBE analyzes competency data to identify trends in learning performance. Such information can help faculty and administrators make decisions about curriculum adjustments and support for student support. CBE can maintain a system of digital grades or ratings based on the competencies achieved by students, which can promote motivation in learning.
AOD approaches use IA to create more personalized, interactive, and adaptive learning environments. IAs serve as the student's personal guide, monitoring individual strengths and weaknesses and learning preferences. Based on this information, IAs can recommend personalized learning pathways, resources, and activities tailored to each student's needs and learning pace. They support collaborative learning by facilitating group discussions, peer evaluation, and collaborative projects. IAs can help coordinate group activities and provide assistance when needed. They automate the grading of assignments, tests and exams, providing instant feedback to students. IAs can also identify common misconceptions and suggest actions to correct them. Administrative IAs assist faculty in course management, scheduling, and monitoring student progress. This allows teachers to focus more on teaching rather than on administrative tasks. IAs can create virtual lab environments and simulations that allow students to conduct experiments or simulations safely and remotely, integrate educational games to provide guidance, assignments, and assessments in a game environment.
It is proposed to implement intelligent learning subsystems through MAS design, which would consist of such components as teacher and student interfaces, an internal message exchange bus, a student guide, a teacher's assistant, a student profile, a supervisor, a KB request agent, an educational resource agent, and an administrator. If necessary, new types of IA with the necessary functions and behavior can be developed and added to the intelligent subsystems, provided that the requirements of the relevant protocols and interfaces are met.
The practical deployment of the system is planned using container technologies and tools. The messaging system [16], KB repositories [18], web servers, web-api services, agent modules are executed under the management of container management systems such as Docker. This architecture provides the possibility of placing the system in various environments, including cloud, flexible scaling, load balancing, optimization of used computing resources. When developing agent modules, it will be possible to use a wide range of programming languages, packages, libraries, frameworks, ensuring their interaction through unified protocols and interfaces and a messaging system.
Summarizing the results, it can be stated that the researched and proposed approaches to the application of KBIA in the field of EL have an important innovative value and actual practical value and can be used to create a real software project and product that will be the development of existing EL systems towards greater interactivity, adaptability and intelligence.