Programma della conferenza

In recent years, the topic of artificial intelligence (AI) has gained increasing importance in research and applications, also in the educational and didactic sphere. One of the aspects investigated in this field is the evaluation of the impact of devices incorporating AI components on learning processes and inclusion, within innovative educational contexts. Analysing the case of conversational chatbots, the report will highlight how such devices are still not designed for interaction with pupils with special educational needs, and even more so with disabilities, excluding them from use and in fact constituting a barrier from an inclusive perspective. In order to intervene on this aspect, IRCIT(International Research Center for Inclusion and Teacher Training) and the company Talent (a training organisation accredited by MIUR and active for years in the field of technological research) have undertaken a collaboration aimed at creating a conversational chatbot capable of interacting with students with all types of functioning: through specific prompt engineering techniques applied to the system message and based on the ICF coding found in school documentation, IRCIT and Talent are working on the development of AskLea, a chatbot capable of adapting communicated imagery, ways of expressing oneself, types of concepts, vocabulary, syntax and emotional tone to the functioning characteristics of different disabilities. In this way, this tool can be used within inclusive and special education, to allow pupils with disabilities to use it directly - and not through the support teacher - in the teaching-learning processes addressed to them.

The TOLC (Test Online CISIA) is an online test used by Italian universities to assess the readiness of prospective students for university studies. It's a crucial gateway to higher education in Italy, currently presents a barrier for blind and visually impaired students. This online assessment, delivered through multiple-choice questions, lacks accessibility features that hinder equal participation. The test is administered electronically, and the questions are presented in a multiple-choice format. However, the TOLC test and its preparatory materials are currently inaccessible to blind and visually impaired students. This paper will discuss the importance of accessibility for blind and visually impaired students, and will outline the specific challenges faced by these students when taking the TOLC test. The paper will also propose a number of recommendations for making the TOLC test and its preparatory materials more accessible.

The project "EFFIMERA – BEYOND TIME" aims to utilize innovative technologies such as virtual reality, augmented reality, and artificial intelligence to make temporary exhibitions of contemporary art from the 20th century accessible and inclusive. Starting from the historical context of art exhibitions, the project aims to explore and reconstruct exhibitions from the second half of the 20th century, no longer accessible, such as Concetto Pozzati's first exhibition in 1968. Through research and documentation of existing materials, interviews, and the creation of multimedia content, an accessible web app will be developed for both computers and smartphones, allowing users to virtually explore past exhibitions enriched with interactive, educational, and experiential features. The project aims to serve as a research tool for the history of exhibitions, revive past exhibitions, create connections between museum institutions, make obtained content available, and enable accessible, educational, and engaging art experiences. Project phases include research and documentation, technology and framework development, and testing and evaluation. Technologies used include virtual reality, augmented reality, and artificial intelligence through the implementation of a web app.

Observation is an essential tool for educators. However, when the observation involves many children in the same context, noises could take the attention from the behaviours to observe. A novel movement in the educational field has emerged that pushes for data-driven action. Especially with the youngest children, challenges arise when collecting quantitative data on social interactions. Current ICT devices allow to obtain large amounts of data from the educational context, and those, along with suitable methods of analysis, can allow innovation.

This systematic review aims to analyse studies that can enable the automatic implementation of social network analysis (SNA) in educational settings. On January 17th, 2024, four databases were inquired. The inclusion criteria were: (1) automatic tool for data collection of social interaction; (2) simultaneous involvement of at least two children; (3) educational setting; (4) age range 0-12 years; (5) type of publication: peer-reviewed journal; (6) language: English; 7) research studies.

We found 21 results that matched the purpose of the study. The explored topics included mapping relationships, social interaction among children with atypical development, studying the spread of infectious diseases, examining homophily, and identifying play types. Only one of the studies was experimental and applied an intervention to promote inclusion; the others were observational or descriptive.

In sum, although various technologies and methods have been validated for automatic social network analysis, we emphasise the lack of application in psycho-educational contexts. Quantitative social interaction data can help professionals to promote quality, engagement, and inclusion through targeted educational action.