Akademik Veri Yönetim Sistemi
5. INTERNATIONAL CONFERENCE ON EDUCATIONAL TECHNOLOGY CONFERENCE AND ONLINE LEARNING, Balıkesir, Türkiye, 26 - 29 Ağustos 2025, ss.81, (Özet Bildiri)
The integration of artificial intelligence (AI) in education has contributed to the development of innovative tools that support personalised learning processes. This paper introduces an AIpowered chatbot that aims to guide students by extracting relevant information from uploaded PDF documents. The aim is to demonstrate the potential of AI in education by accelerating access to information, enhancing engagement with academic resources, and supporting independent learning. The design adopts a Retrieval-Augmented Generation (RAG) approach, which aims to improve the accuracy of responses by combining document retrieval, text processing, and large language modelling (LLM) components. The primary objective of the research is to enable students to access academic resources and to provide rapid and precise access to information. The system has been developed using LLM and tested on educational materials in different disciplines. The chatbot processes user queries in English or Turkish, identifies the most relevant document, highlights the relevant text on the screen, and provides context-appropriate answers. The study's findings suggest that the chatbot has the potential to increase engagement with academic resources by reducing search time and thus support learning efficiency. However, user feedback indicates that there is room for improvement with regard to accessibility and ease of use, particularly in independent learning environments. In addition, issues such as the management of ambiguous queries and ensuring the accuracy of extracted information are considered as areas for future development. In the future, improvements to the chatbot's contextual interpretation capabilities, expansion of its language support, and enrichment of its multimedia content integration are planned in line with user feedback.