Language, Speech and Interaction
Our main areas of interest are speech, vision and language processing, machine learning and interaction.
The purpose of our research is to investigate how humans process speech, language and vision, and to design mathematical models for automatic processing that can be used with communicating machines.
Our studies rely on computational models from batch or on-line examples and provide predictive models for data classification, automatic sequence labelling and structure inference.
Another research area the group is involved in is human-centered communication systems. We investigate models of interactions in any ICT scenario, such as speech-to-speech, speech-to-web and multimodal interactions. Our focus is both on the computational aspect of the interaction models as well as on the usability of user interfaces.
Finally, we also devote our attention to collaborative systems and web architecture. We are active in the design of software architecture for advanced web and multimedia collaborative systems.
Publications
15 publications for 5 currently enrolled students
Response to: “Converging Approaches to Autistic Online Discourse”
Carollo, Alessandro; Fong, Seraphina; Vivanti, Giacomo; Messinger, Daniel S.; Dimitriou, Dagmara; Esposito, Gianluca
in AUTISM RESEARCH, v. 19, n. 2 (e70187) (2026). - Publication URL . - DOI: 10.1002/aur.70187
Accuracy of Autism-Related TikTok Information in Italian: A Comparison Between Human Raters and Large Language Models
Carollo, Alessandro; Fong, Seraphina; Belardinelli, Giovanni; Perzolli, Silvia; Vivanti, Giacomo; Messinger, Daniel S.; Dimitriou, Dagmara; Esposito, Gianluca
in JOURNAL OF AUTISM AND DEVELOPMENTAL DISORDERS, v. 2026, (2026). - Publication URL . - DOI: 10.1007/s10803-026-07249-9
CrisiText: A dataset of warning messages for LLM training in emergency communication
Gonella, Giacomo; Campedelli, Gian Maria; Menini, Stefano; Guerini, Marco
in Findings of the Association for Computational Linguistics: EACL 2026, Rabat, Morocco: ACL - Association for Computational Linguistics, 2026, p. 6657-6677. - ISBN: 979-8-89176-386-9. Proceedings of: EACL, Rabat, February 2026. - Publication URL . - DOI: 10.18653/v1/2026.findings-eacl.350
LLMs as Repositories of Factual Knowledge: Limitations and Solutions
Mousavi, S. M.; Alghisi, S.; Riccardi, G.
in IEEE TRANSACTIONS ON AUDIO, SPEECH, AND LANGUAGE PROCESSING, v. 34, (2026), p. 2213-2226. - DOI: 10.1109/TASLPRO.2026.3680709
Autism Spectrum Disorders Discourse on Social Media Platforms: A Topic Modeling Study of Reddit Posts
Fong, Seraphina; Carollo, Alessandro; Vivanti, Giacomo; Messinger, Daniel S.; Dimitriou, Dagmara; Esposito, Gianluca
in AUTISM RESEARCH, v. 18, n. 8 (2025), p. 1608-1619. - Publication URL . - DOI: 10.1002/aur.70066
Speech LLMs in Low-Resource Scenarios: Data Volume Requirements and the Impact of Pretraining on High-Resource Languages
Fong, Seraphina; Matassoni, Marco; Brutti, Alessio
in Proceedings of Interspeech, ISCA - International Speech Communication Association: ISCA - International Speech Communication Association, 2025, p. 2003-2007. - (INTERSPEECH). Proceedings of: 26th Interspeech Conference 2025, Rotterdam, The Netherlands, 17-21 August 2025. - Publication URL . - DOI: 10.21437/Interspeech.2025-764
Dynamic Topic Modeling of Kratom Use and Experiences: Insights on 13 Years of Reddit Discussions
Fong, Seraphina; Carollo, Alessandro; Prevete, Elisabeth; Corazza, Ornella; Esposito, Gianluca
in INTERNATIONAL JOURNAL OF MENTAL HEALTH AND ADDICTION, v. 2025, (2025). - Publication URL . - DOI: 10.1007/s11469-025-01596-x
Characterizing motion artifacts in functional near-infrared spectroscopy signals using ground-truth movement information and computer vision
Bizzego, Andrea; Carollo, Alessandro; Fong, Seraphina; Furlanello, Cesare; Esposito, Gianluca
in BIOMEDICAL SIGNAL PROCESSING AND CONTROL (ONLINE), v. 110, Part B, n. December 2025, 108256 (2025). - Publication URL . - DOI: 10.1016/j.bspc.2025.108256
CIVET: Systematic Evaluation of Understanding in VLMs
Rizzoli, Massimo; Alghisi, Simone; Khomyn, Olha; Roccabruna, Gabriel; Mousavi, Seyed Mahed; Riccardi, Giuseppe
in Findings of the Association for Computational Linguistics: EMNLP 2025, 209 N. Eighth Street, Stroudsburg, PA, USA, 18360: Association for Computational Linguistics (ACL), 2025, p. 4462-4480. - ISBN: 979-8-89176-335-7. Proceedings of: 30th Conference on Empirical Methods in Natural Language Processing, EMNLP 2025, Suzhou, China, 4th November-9th November 2025. - Publication URL . - DOI: 10.18653/v1/2025.findings-emnlp.239
Ozempic (Glucagon-like peptide 1 receptor agonist) in social media posts: Unveiling user perspectives through Reddit topic modeling
Fong, Seraphina; Carollo, Alessandro; Lazuras, Lambros; Corazza, Ornella; Esposito, Gianluca
in EMERGING TRENDS IN DRUGS, ADDICTIONS, AND HEALTH, v. 4, n. 100157 (2024). - Publication URL . - DOI: 10.1016/j.etdah.2024.100157
Captagon: A comprehensive bibliometric analysis (1962–2024) of its global impact, health and mortality risks
Fong, Seraphina; Carollo, Alessandro; Rossato, Andrea; Prevete, Elisabeth; Esposito, Gianluca; Corazza, Ornella
in SAUDI PHARMACEUTICAL JOURNAL, v. 32, n. 11 (2024). - Publication URL . - DOI: 10.1016/j.jsps.2024.102188
Are LLMs Robust for Spoken Dialogues?
Mousavi, Seyed Mahed; Roccabruna, Gabriel; Alghisi, Simone; Rizzoli, Massimo; Ravanelli, Mirco; Riccardi, Giuseppe
in Proceedings of the 14th International Workshop on Spoken Dialogue Systems Technology, sapporo, Japan: 14th International Workshop on Spoken Dialogue Systems Technology, 2024. Proceedings of: IWSDS2024, JAPAN, 04/03/2024
Computer Vision-Driven Movement Annotations to Advance fNIRS Pre-Processing Algorithms
Bizzego, Andrea; Carollo, Alessandro; Senay, Burak; Fong, Seraphina; Furlanello, Cesare; Esposito, Gianluca
in SENSORS, v. 24, n. 21 (2024), p. 6821. - Publication URL . - DOI: 10.3390/s24216821
Should We Fine-Tune or RAG? Evaluating Different Techniques to Adapt LLMs for Dialogue
Alghisi, Simone; Rizzoli, Massimo; Roccabruna, Gabriel; Mousavi, Seyed Mahed; Riccardi, Giuseppe
in Proceedings of the 17th International Natural Language Generation Conference, Association for Computational Linguistics: Association for Computational Linguistics (ACL), 2024, p. 180-197. - ISBN: 9798891761223. Proceedings of: 17th International Natural Language Generation Conference, INLG 2024, Tokyo, Japan, 23/09/2024 - 27/09/2024. - Publication URL . - DOI: 10.18653/v1/2024.inlg-main.15
DyKnow: Dynamically Verifying Time-Sensitive Factual Knowledge in LLMs
Mousavi, Seyed Mahed; Alghisi, Simone; Riccardi, Giuseppe
in Findings of the Association for Computational Linguistics: EMNLP 2024, Miami, Florida, USA: Association for Computational Linguistics, 2024, p. 8014-8029. Proceedings of: EMNLP2024, Miami, Florida, USA, november 2024. - Publication URL
