This paper describes a method and technology for processing natural language texts and extracting data from the text that correspond to the semantics of an ontological model. The proposed method is distinguished by the use of a Large Language Model algorithm for text analysis. The extracted data are stored in an intermediate format, after which individuals and properties that reflect the specified semantics are programmatically created in the ontology. The proposed technology is implemented using the example of an ontological model that describes the geographical configuration and administrative–territorial division of Kazakhstan. The proposed method and technology can be applied in any subject areas for which ontological models have been developed. The results of the study can significantly improve the efficiency of using knowledge bases based on semantic networks by converting texts in natural languages into semantically linked data.

The emotional speech recognition method presented in this article was applied to recognize the emotions of students during online exams in distance learning due to COVID-19. The purpose of this method is to recognize emotions in spoken speech through the knowledge base of emotionally charged words, which are stored as a code book. The method analyzes human speech for the presence of emotions. To assess the quality of the method, an experiment was conducted for 420 audio recordings. The accuracy of the proposed method is 79.7% for the Kazakh language. The method can be used for different languages and consists of the following tasks: capturing a signal, detecting speech in it, recognizing speech words in a simplified transcription, determining word boundaries, comparing a simplified transcription with a code book, and constructing a hypothesis about the degree of speech emotionality. In case of the presence of emotions, there occurs complete recognition of words and definitions of emotions in speech. The advantage of this method is the possibility of its widespread use since it is not demanding on computational resources. The described method can be applied when there is a need to recognize positive and negative emotions in a crowd, in public transport, schools, universities, etc. The experiment carried out has shown the effectiveness of this method. The results obtained will make it possible in the future to develop devices that begin to record and recognize a speech signal, for example, in the case of detecting negative emotions in sounding speech and, if necessary, transmitting a message about potential threats or riots.

This article is dedicated to the development of a model for competencies within an educational program and its implementation through the use of semantic technologies. The model proposed by the authors is distinctive in that competencies are organized into a hierarchical data structure with arbitrary levels of nesting. Furthermore, the article presents an original solution for modelling the input requirements for studying a course, which is defined in the form of dependencies between the competencies generated by the course and the competencies of other courses. The outcome of this work is an ontological model of a competency-based curriculum, for which the authors have developed and implemented algorithms for data addition and retrieval, as well as for analyzing the consistency of the curriculum in terms of the input requirements for studying a discipline and the learning outcomes from previous periods. The findings presented in the article will prove to be valuable in the development of educational process management information systems and educational program constructors. They will also be instrumental in aligning diverse educational programs within the context of academic mobility.

The article presents the results of the application of an ontological approach to the description of the structure and content of the educational programme, and its subsequent analysis for the consistency of prerequisites and learning outcomes of courses. The practical result of the work is an ontology approach implemented in the Protégé 5.5.0 editor, which reflects the studied disciplines in terms of the skills they form and the entrance requirements (prerequisites) for the qualification of the student. The curriculum model includes sequences of semesters and courses of study (academic year) related by time relationships. The developed ontology approach is filled with data from the educational programme “Software Engineering”. The authors have earned queries in DL Query and SPARQL languages, which, using logical inference procedures, make it possible to analyse an educational programme for consistency of disciplines in terms of input requirements and the skills of the learner formed during the training period. The developed ontology and rules of logical inference can be used as a part of the educational process management information systems and educational programme designers, for the intellectual analysis of programme integrity and the consistency of learning prerequisites and outcomes in disciplines.

The problem of multiple zones in computer vision, including pattern recognition in the agricultural sector, occupies a special place in the field of artificial intelligence in the modern aspect. The object of the study is the recognition of weeds based on deep learning and computer vision. The subject of the study is the effective use of neural network models in training, involving classification and processing using datasets of plants and weeds. The relevance of the study lies in the demand of the modern world in the use of new information technologies in industrial agriculture, which contributes to improving the efficiency of agro-industrial complexes. The interest of private agricultural enterprises and the state is caused by an increase in the yield of agricultural products. To recognize weeds, machine learning methods, in particular neural networks, were used. The process of weed recognition is described using the Mark model, as a result of processing 1,562 pictures, segmented images are obtained. Due to the annual increase in weeds on the territory of Kazakhstan and in the course of solving these problems, a new plant recognition code was developed and written in the scanner software module. The scanner, in turn, provides automatic detection of weeds. Based on the results of a trained neural network based on the MaskRCNN neural network model written in the scanner software module meeting new time standards, the automated plant scanning and recognition system was improved. The weed was recognized in an average of 0.2 seconds with an accuracy of 89 %, while the additional human factor was completely removed. The use of new technology helps to control weeds and contributes to solving the problem of controlling them Проблема множинності зон у комп'ютерному зорі, зокрема розпізнавання образів в агросекторі займає особливе місце в області штучного інтелекту у сучасному аспекті. Об'єктом дослідження є розпізнавання бур'янових рослин на основі глибокого навчання і комп'ютерного зору. Предметом дослідження є ефективне використання моделей нейронних мереж у навчанні, що включає в себе класифікацію та обробку з використанням дата сетів рослин та бур'янів. Актуальність дослідження полягає у вимозі сучасного світу щодо застосування нових інформаційних технологій в агропромисловій сфері, що сприяє підвищенню ефективності роботи агропромислових комплексів. Інтерес приватних сільгосппідприємств та держави викликаний збільшенням зростання врожайності сільгосппродуктів. Для розпізнавання бур'янів використовувалися методи машинного навчання, зокрема нейронні мережі. За допомогою моделі Марка описаний процес розпізнавання бур'янів, в результаті обробки 1562 картин отримані сегментовані зображення. У зв'язку зі щорічним посиленням бур'янів на території Казахстану і в ході вирішення цих проблем розроблений новий код для розпізнавання рослин та прописаний в програмний модуль сканера. Сканер, в свою чергу, забезпечує автоматичне виявлення бур'янів. За результатами навченої нейронної мережі на основі моделі нейромережі MaskRCNN, прописаної в програмному модулі сканера, що відповідає новим стандартам часу, покращена автоматизована система сканування та розпізнавання рослин. Бур'янову рослину було розпізнано в середньому за 0,2 секунди з показником точності 89 %, при цьому додатковий людський фактор був повністю видалений. Застосування нових технологій допомагає боротися з бур'янами та сприяє вирішенню проблеми боротьби з ними