Svetlana B. Shmakova

Artificial intelligence technologies open up new prospects for the education system, however, limited understanding of the diversity of neural networks and insufficient training of teachers to use them slow down this process. In the context of the digital educational environment, visualization of educational information is of particular importance due to the increase in the volume of educational material and the peculiarities of perception, memorization and reproduction of information by modern schoolchildren. Understanding the capabilities of neural network services for developing visual content is also important in order to offer teachers an additional tool when designing a lesson based on them. An electronic lesson constructor can be such a tool. The aim of the study is a comparative analysis of the use of neural networks contained in the electronic lesson constructor "Best Digital Lesson" to create educational visualization. The main method is a survey of teachers of the Udmurt Republic, which provided data on the real practice of using neural networks for educational visualization. The article reveals the potential of practical application of the electronic lesson constructor "The Best Digital Lesson", in particular the section "Digital Wheel", which contains a list of neural networks designed to create educational visualization. The listed resources are distributed according to three levels of students' mental activity: memorization (Kandinsky, Shedevroom, Ideogram AI), understanding (yandex GPT, GigaChat, ChatGPT, Rytr, litmaps, perplexity.ai, beta.diffit.me, chefdroyd) and analysis (app.yoodli.ai, app.songr.ai, gliglish.com, murf.ai). A review of the listed tools is provided, their target systematization is proposed. During the analysis, all the considered neural networks are assessed from the point of view of their applicability for creating visual materials in the process of designing a lesson. The theoretical significance of the work consists in generalizing the results of domestic and foreign studies on the issue of using neural networks in education, in in-depth understanding of the potentials of their application for creating visualization in educational purposes. The practical significance is determined by the development of methodological recommendations for the use of neural networks in the process of designing and conducting classes with elements of visualization based on the work of teachers with an electronic lesson constructor.