Programma della conferenza

The latest frontier in Intelligent Adaptive Learning Systems (IALS) is conversational intelligent systems, more commonly known as chatbots; among them, the most popular is ChatGPT, which we used in this research to support designing inclusive mathematics story-problems. Research was conducted with 175 prospective secondary school special education teachers who attend the “Special Education Laboratory: codes of logical and mathematical language” as part of the Specialization Course for Special Education Teachers at the University of Molise. The research questions are: (i) How is the support offered by ChatGPT perceived by the prospective special education teachers? (ii) Can the use ChatGPT has a positive impact on the design of inclusive teaching activities in mathematics? (iii) What are the pitfalls of using ChatGPT as a support in designing learning activities in mathematics by prospective special education teachers? Participants worked in small groups and were asked to design inclusive mathematics story-problems. It was important to carry out this task first without using ChatGPT and then by using it. At the end of the course, participants were asked to answer a semi-structured self-directed questionnaire using CAWI methodology. Preliminary analysis seems to show that ChatGPT was perceived by participants as a useful tool that can free up energy to be allocated to higher levels of creative production by speeding up the time. In addition, it was necessary to carry out checks by participants on the correctness of what ChatGPT proposed from a mathematical point of view.

The issuance of the Operating Instructions on "STEM and Multilingual Skills Enhancement Actions" (Ministerial Decree 65/2023) directs the long term and daily work of teachers towards active teaching linked to disciplinary content, within a STEAM-type model. The use of AI can effectively support the implementation of educational, training and guidance paths for students within the curricula of all school cycles of activities. AI is particularly useful in promoting the integration of activities, methodologies and content aimed at developing STEM, digital and innovation skills. In addition, artificial intelligences, particularly generative one, are powerful tools with pros and cons relevant to the world of learning at school. They may strongly support integration of skills.

This paper presents an interactive Project implemented on LMS Moodle as part of a Scientific High School's Digital Citizenship Education.

The teaching of physics in some scientific degree courses does not include laboratory activities or includes very few of them. Laboratory experimental activities are essential because they enhance the understanding and learning of fundamental concepts and laws that describe scientific phenomena. To implement these experiential activities, we are developing a serious game in Unreal Engine, focused on Physics, and based on the concept of an escape room. This game aims to facilitate the understanding of the concepts by using simulations, ensuring a learning experience amplified by engagement, curiosity, interaction, and feedback. Within the game, players are guided by a narrative path which leads them through different rooms, each presenting puzzles and challenges. By solving these puzzles, players progress in the game. This method aligns with research on "Game-Based Learning and Serious Games in Education," which highlights the importance and effectiveness of game-based learning in modern teaching. A fundamental role in the game is played by an intelligent tutor based on AI, which actively assists players during the adventure. This system, which learns from questions posed by students, progressively specializes, offering personalized and targeted educational support. Thanks to this AI guidance, students receive immediate assistance on specific issues, further enhancing their understanding of the topics addressed. The goal of the project is to evaluate the game’s impact on student learning. This is achieved by implementation of pre-tests and post-tests. Both are designed to measure not only the theoretical knowledge acquired by students, but also the practical skills developed through the game.

Continuous training for company employees is evolving in an increasingly important and continuous way. There are many reasons and they have to do both with the rapid obsolescence of knowledge and with the need for transversal skills that cross specialist knowledge. All this must happen compatibly with work activities, deadlines, organizational constraints of space and time. In this sense, continuous training for employees represents fertile ground for experimentation and innovation of forms and tools. The paper describes the implementation of a generative AI-based virtual tutor within the Moodle platform to enhance the educational experience. The ongoing experimentation at a company that carries out online training for sales staff is structured in three phases: 1) integration of AI into the Learning Management System; 2) training on specific course content; 3) integration of a virtual avatar for voice interactions. The virtual tutor uses GPT-4 to provide: (a) personalized support 24/7, reinforcing learning through immediate feedback, (b) insights on specific topics, (c) interactive quizzes, (d) multilingual support and potential integration of external resources into course content. The effectiveness of the assistant will be assessed based on the collection and analysis of student feedback, with ongoing adjustments and improvements. Finally, the document emphasizes the importance of privacy and suggests the evaluation of open-source LLMs for enhanced data protection. Moreover, the importance of training through Retrieval Augmented Generation (RAG) is highlighted, and some tips are provided to ensure targeted and personalized responses.