- Спірін, О.М. (orcid.org/0000-0002-9594-6602), Коломієць, А. (orcid.org/0000-0003-0536-0147), Громов, Є. (orcid.org/0000-0002-0234-606X), Жовнич, О. (orcid.org/0000-0001-6430-7341), Коломієць, Д. (orcid.org/0000-0003-1966-0837) and Кушнір, О. (orcid.org/0009-0002-6254-6589) (2025) Use of the Deep Research AI tool in educational and scientific-pedagogical activities Інформаційні технології і засоби навчання, 6 (110). pp. 271-293. ISSN 2076-8184
![[thumbnail of 2446.pdf]](https://lib.iitta.gov.ua/style/images/fileicons/text.png "2446.pdf") |
Text
2446.pdf Download (1MB) |
Abstract
The article is devoted to the study of the essence and functionality of “Deep Research AI” as an artificial intelligence tool, which is currently actively used in its various models. Such functionality, which is implemented as an additional mode or tool in various chatbots/AI systems,is aimed at automated collection, detailed analysis and synthesis of complex information, in particular scientific information, from available sources. The article presents in detail the principles of operation of such an AI tool, its advantages and limitations, as well as potential scenarios for its application in scientific, pedagogical and scientific-pedagogical activities. Special attention is paid to how this AI tool can contribute to supporting scientific research of teachers and developing critical thinking of students. It is demonstrated how the use of Deep Research AI makes significant changes in the field of scientific and pedagogical research, contributes to the generation of new knowledge and innovations in education. The authors compared the quality of reports of different AI models generated using the Deep Research AI tool for the same query regarding the possibilities of using AI in STEAM education. The following indicators were subject to expert evaluation: the volume of the received text, the exhaustiveness of the response structure (the number of aspects of the query problem taken into account), the correctness, clarity of the response, the number of sources processed, the relevance of the list of proposed literature. The role of experts was played by 5 doctors of pedagogical sciences, who analyzed the reports of 11 AI models. According to the results of expert evaluation and comparative analysis, the highest average score was given to the Gemini Advanced model, which prepared a 37-page report in the form of a scientific article with all the necessary structural elements (abstract, keywords, introduction, theoretical foundations, review of existing research, examples of AI applications in STEAM education, advantages, challenges, limitations, conclusions and recommendations) and presented a list of 69 sources used, each of which related to the issue of the query. It was concluded that the Deep Research AI tool will be able to adapt to the needs of a specific user (teacher, student, researcher), offering individual report formats. The uniqueness of Deep Research AI as an assistant in pedagogical research was demonstrated, and the need for its responsible use by scientists and education seekers was emphasized.
Downloads
Downloads per month over past year
Actions (login required)
 |
View Item |