Month: July 2024

Corpora Creation for the Extraction of Metalinguistic Concepts

The project aims to research the most commonly used metalinguistic notions in the French language, especially at the lessical level.

The project involves the creation of a corpus of textbooks, grammars, essays, and lexicographical material, the extraction of metalinguistic notions, and the identification of a possible progression applicable in educational contexts.

This progression will be tested among different groups of learners:
a) adult learners who are Italian speakers with work-related needs in business settings;
b) adult learners with a first language (L1) other than Italian, who are partially Italian-speaking;
c) adult learners who are non-Italian speakers;
d) adult learners specializing in the communication sector;
e) learners with special educational needs (SEN).

Group leader: Maria Francesca Bonadonna

Internal members:

  • Alessia Della Rocca
  • Francesco Faresin (PhD student)
  • Paolo Frassi
  • Giovanni Luca Tallarico

External member:

  • Fatna Cherif Hosni

 

Chinese Linguistics with an Inclusive Perspective

The project aims to explore the accessibility and inclusiveness of Mandarin Chinese texts related to institutions and health, including both physical and emotional health. The project focuses on inclusive language for the Taiwanese LGBTQIA+ community and on the communicative differences between institutions and private individuals, as well as the emotional well-being of pet owners who have lost their pets.

The project involves the creation of at least two corpora of texts, which will be analyzed using corpus-based approaches, discourse analysis, cognitive linguistic approaches such as Conceptual Metaphor Theory, and quantitative analyses. Finally, guidelines/manuals for language will be prepared in light of the analyses conducted, to ensure they are easily accessible to researchers and/or the non-specialist public.

Group leader: Michele Mannoni

Internal members:

  • Marco Casentini (postdoctoral researcher) 

The fourth episode of Inclusioni is online

The fourth episode of Inclusioni, the podcast channel of the Department of Foreign Languages and Literatures in collaboration with Fuori Aula Network, is now available on major podcast platforms. In this episode, the hosts embark on a journey to Australia, exploring a past that is often overlooked or even denied, and a present where the legacy of colonialism remains very strong.

Did you know that Australian Aboriginal culture is the oldest in the world, with traces dating back over 60,000 years? What is the current condition of Indigenous Australians? What was voted on in the 2023 constitutional referendum?

Our guide for this episode is Joaquin Rudd, an Australian singer-songwriter, musician, and activist. Joaquin was also a guest at the event “Songlines: the ancient secret to our planet’s oldest surviving culture”, held on March 25 at Santa Marta. Along with Fred Leone, he discussed and set to music the sounds of Australian Aboriginal tribal cultures.

The interview is conducted by Chiara Battisti, a professor of Anglo-American Languages and Literatures at the Department of Foreign Languages and Literatures – University of Verona.

Happy listening!

Multilingualism and Linguistic Inclusion in Post-Migration Society at Viadrina

Primo piano del palco presso la sala del Decanato della Europa-Universität Viadrina, dove siedono alle poltrone quattro interlocutori

On Wednesday, July 3, the presentation of the book “Sprachwechsel – Perspektivenwechsel? Mehrsprachigkeit und kulturelle Vielstimmigkeit in der deutschsprachigen Gegenwartsliteratur”, edited by Gabriella Pelloni and Ievgeniia Voloshchuk, took place in the Dean’s Hall of the Europa-Universität Viadrina. This event is part of a long-standing collaboration between the Department of Foreign Languages and Literatures of the University of Verona and the Faculty of Cultural Studies at Viadrina University. The subsequent debate touched on topics such as multilingualism and linguistic inclusion in the current post-migration society.

Great Success for the First Food Talk

Inizio del primo Food Talk in cattedra Fiorato, Battisti e il relatore Anmol Deep Singh

The first Food Talk of the “Inclusive Humanities” Project of Excellence, titled “Food and Cultural Identity in Anita Desai’s Fasting, Feasting”, was held on Tuesday, 2 July, at the Santa Marta campus. The talk, given by Dr. Anmol Deep Singh, is the first in a series of Food Talks conceived and curated by Chiara Battisti and Sidia Fiorato that explore the role of food and the typicality of different gastronomic cultures in world literature novels and are open to the public to promote dialogue and social inclusion.

Anmol Deep Singh offered reflections and comparisons between Indian and U.S. cultures through food, the act of eating, and sharing, analysing the novel “Feasting, Fasting” by Anita Desai. The narrative’s timeline coincides with the time it takes for a tea package, shipped from India by older sister Uma, to reach her younger sister Aruna in the United States. Thousands of miles apart, the two sisters live completely different lives: two different cultures reflected through food. Uma, the older sister, burdened with a large and varied family, still cannot leave the nest and looks disapprovingly at Aruna, the younger sister, who, having fled abroad for love, goes on to build a perfect family. The narrative then shifts to Massachusetts, America, where Aruna’s son observes the Patton family’s bewildering life, caught between meat binges and deprivation bordering on anorexia. The novel compares two different worlds that drive the characters to renunciation or excess, to fasting or feasting: the tight-knit, suffocating Indian family and the cold, indifferent freedom American household, where life appears free from rules and obligations.

The talk finally concluded with a themed aperitif offered to the participants.

Inclusion and Computer Science

Il progetto analizza l’impatto dell’Intelligenza Artificiale (IA) e del Natural Language Processing (NLP) sui temi della diversità, dell’equità e dell’inclusione, un ambito di crescente rilevanza nella ricerca contemporanea in informatica e intelligenza artificiale.

Da un lato, il progetto esamina le criticità che le tecnologie di IA possono introdurre o accentuare in termini di inclusione. Ciò comprende l’analisi di bias sociali e culturali—ad esempio legati al genere, alla razza, alla religione o alla disabilità—presenti nei modelli computazionali addestrati su dataset di grandi dimensioni (Bombieri et al., 2025a; Bombieri et al., 2025b; Bombieri & Rospocher, 2025).

Dall’altro lato, il progetto si concentra sullo sviluppo, la sperimentazione e l’implementazione di strumenti, metodi, tecnologie e modelli computazionali concreti finalizzati a promuovere l’inclusione e a migliorare l’accessibilità dei contenuti. In particolare:

  • studiamo e sperimentiamo metodologie per migliorare l’accessibilità dei materiali didattici digitali (Boschiero et al., 2024);
  • sviluppiamo strumenti e approcci (ad esempio screen reader e sistemi di programmazione vocale) per migliorare l’accessibilità alla programmazione, come il Python Code Reader;
  • contribuiamo al progetto interno AI4EDU dell’Università di Verona, coordinato dal gruppo Tecnologie Innovative per la Didattica (Ufficio IT e Comunicazione), che ha condotto uno studio esplorativo sull’uso delle tecnologie di IA all’interno della piattaforma di e-learning dell’Ateneo (Moodle).

In relazione a queste linee di ricerca, curiamo come guest editor il numero speciale The Role of Artificial Intelligence for Diversity, Equity, and Inclusion (Information, MDPI). La call for papers è attualmente aperta, con la scadenza per l’invio dei contributi fissata al 31 ottobre 2026.

Il progetto è inoltre dedicato allo studio dei bias nei confronti dell’Intelligenza Artificiale stessa. In collaborazione con Numero Cromatico, analizziamo le reazioni e le percezioni dei lettori rispetto a narrazioni prodotte da modelli linguistici generativi. L’obiettivo è comprendere come tali testi vengano valutati in termini di gradimento, coinvolgimento emotivo, valore artistico e inclusività. Da notare, come osservato sperimentalmente, che anche i modelli di IA manifestano un bias negativo nei confronti dei testi letterari attribuiti ad autori artificiali (Rospocher et al., 2026).

Responsabile del progetto: Marco Rospocher

Membri interni

  • Manuel Boschiero
  • Simone Rebora
  • Massimo Salgaro

Membri esterni

  • Marco Bombieri (Università di Trento)
  • Olga Forlani, Silvano Pasquali (Direzione ITC, Università di Verona)
  • Davide Quaglia (Dipartimento di Informatica, Università di Verona)
  • Numero Cromatico

Azioni: WP1.13

Riferimenti bibliografici

Manuel Boschiero, Marco Rospocher, Olga Lucia Forlani and Silvano Pasquali. Migliorare l’accessibilità dei materiali didattici digitali nel contesto universitario: un caso di studioAtti del MoodleMoot Italia 2023, pp. 61–70. MediaTouch 2000 (In co-edizione con Associazione Italiana Utenti Moodle A.p.s (AIUM), Università degli Studi di Firenze). ISBN: 978-88-907493-9-1 (2024)

Marco Bombieri, Simone Paolo Ponzetto and Marco Rospocher. Do LLMs Authentically Represent Affective Experiences of People with Disabilities on Social Media?Proceedings of the Eleventh Italian Conference on Computational Linguistics (CLiC-it 2025), Cagliari, Italy, September 24-26, 2025, vol. 4112 of CEUR Proceedings. CEUR-WS.org (2025)

Marco Bombieri, Simone Paolo Ponzetto and Marco Rospocher. The Dangerous Effects of a Frustratingly Easy LLMs Jailbreak AttackIEEE Access, vol. 13, pp. 126418–126431 (2025)

Marco Bombieri and Marco Rospocher. Mining Impersonification Bias in LLMs via Survey FillingInformation, vol. 16, no. 11. ISSN: 2078-2489 (2025)

Marco Rospocher, Massimo Salgaro and Simone Rebora. Machines Prefer Humans as Literary Authors: Evaluating Authorship Bias in Large Language ModelsInformation, vol. 17, no. 1. ISSN: 2078-2489 (2026)