@inproceedings{nokey,
title = {AI4EDU: An innovative conversational AI assistant for teaching and learning},
author = {Anna Vacalopoulou and Viktor Gardelli and Theodoros Karafyllidis and Foteini Liwicki and Hamam Mokayed and Marios Papaevripidou and Georgios Paraskevopoulos and Spyridoula Stamouli and Athanasios Katsamanis and Vassilis Katsouros},
url = {https://library.iated.org/view/VACALOPOULOU2024AI4},
doi = {10.21125/inted.2024.1877},
isbn = {978-84-09-59215-9},
year = {2024},
date = {2024-03-12},
urldate = {2024-03-12},
booktitle = {INTED2024 Proceedings},
pages = {7119-7127},
publisher = {IATED},
abstract = {Artificial intelligence (AI) is transforming education by introducing innovative tools and approaches to enhance teaching and learning. Chatbots like Dawebot and TA-bot serve as intelligent student assistants, providing interactive quizzes, immediate feedback, and personalized support. AI-based systems like Rexy and Kwame offer personalized learning experiences, adapting to individual student needs and enhancing self-regulated learning. Chatbots and AI-based systems can address educational challenges in resource-constrained settings, providing scalable and cost-effective learning solutions. More specifically, Generative AI technologies, like the Large Language Model-based ChatGPT, with their advanced language capabilities, promise enriched and interactive real-time learning through natural conversation. These advancements demonstrate the potential of AI to revolutionize education, making it more personalized, accessible, and effective for all learners. Some typical ways in which this potential is demonstrated is the ability to quickly answer questions, provide explanations according to preset criteria, and suggest additional teaching and learning resources.
The Conversational AI assistant for teaching and learning (AI4EDU) project aims to improve school education by exploring and implementing new approaches, technologies, and applications of AI in Education. It seeks to create new, innovative, adaptable, personalized, engaging, and effective ways of teaching and learning, in view of the growing challenges brought on by the digital transformation of education. This paper describes the design and development of a student-facing AI platform to enhance, personalize, and facilitate learning, empower teachers, and support teaching and assessment objectives, increase knowledge and understanding of AI among students and teachers to understand how AI applications work, their potential, and their limitations. A key feature of this endeavor is the incorporation of Large Language Models (LLMs) like ChatGPT. Incorporating LLMs into educational environments poses challenges like restricted personalization, the risk of misinformation, and excessive dependence on technology. In order to combine LLMs with textbooks and other educational resources currently used in schools, AI4EDU tackles these issues by employing Retrieval Augmented Generation (RAG) and a conversational AI model for interacting with students and teachers through natural language understanding. Key aspects include ethical considerations, content reliability, and a balance between AI assistance and human interaction. The project seeks to transform education by delivering personalized, adaptive, and captivating learning experiences for both students and teachers.
To combine all these into a flexible, personalized, engaging, and effective educational platform, AI4EDU develops two different tools that accommodate the needs of both students and teachers. These tools are Study Buddy, a conversational AI assistant to help students with a wide range of learning activities, and Teacher Mate, a dashboard and chatbot that also incorporates AI-powered technologies and assists teachers in organizing the teaching process, preparing materials and tests, evaluating students, tracking their progress, and more. Both tools incorporate educational resources from four EU countries (Cyprus, Greece, Ireland, Sweden) and are offered in three European languages (Greek, Irish, Swedish).},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}