PREPARING FUTURE COMPUTER SCIENCE TEACHERS FOR THE INTEGRATION OF ARTIFICIAL INTELLIGENCE INTO THE EDUCATIONAL PROCESS: THE CASE OF A CHATBOT

Abstract

The article presents an example of creating an educational chatbot based on the Retrieval-Augmented Generation (RAG) architecture in the process of training future computer science teachers. The focus is on the practical application of modern artificial intelligence technologies in pedagogical activities and the development of students' digital competencies. The relevance of studying contemporary approaches to implementing AI systems (prompting, fine-tuning, retrieval-only, RAG) by future computer science teachers is also substantiated. A comparative analysis of these approaches was conducted, taking into account their technical and educational feasibility. The technical implementation of the created chatbot is described in detail, from document uploading and semantic search based on embedding vectorization to the generation of informative responses using a GPT model, which ensures interactive student engagement. The article presents the results of a survey of students majoring in "Secondary Education (Computer Science)," which demonstrated a high level of interest and the relevance of the chosen research topic, as well as the necessity of developing a training course on the use of AI by future computer science teachers. This course would cover not only the user-oriented approach to AI utilization but also a deeper approach through the design of AI systems and interfaces for them. Additionally, the educational advantages and limitations of implementing RAG technologies in the pedagogical practice of training future computer science teachers are highlighted. The authors emphasize the pedagogical value of the RAG chatbot as a tool that facilitates the thoughtful application of AI in teaching and enhances the ability of future computer science teachers to independently design and use AI bots in educational activities.