2022 | Italy | Promising | Schools/Education | SDG10 | SDG3
BullyBuster – A framework for bullying and cyberbullying action detection by computer vision and artificial intelligence methods and algorithms

Company or Institution

"BullyBuster" team made up of researchers from four italian academies: university of Bari, Cagliari, Foggia, Napoli "Federico II"







Sustainable Development Goals (SDGs)

SDG 3: Good Health and Well-being

SDG 10: Reduced Inequality

General description of the AI solution

In this project, four RUs from South Italy’s Academies synergically cooperate to develop a set of artificial intelligence-based software tools for bullying and cyberbullying actions detection. Their proposal starts from the state-of-the-art on behavioural biometrics and crowd analysis in computer vision systems and the recent efforts of criminal behavioural modelling in psychology and juridical fields. The tools are operating on different kinds of data sources: (1) video-based analysis, by segmentation and haracterization of the scene by means of temporal and spatial textural descriptors, in order to detect specific bullying actions on the basis of the crowd movements around the victim and, where possible, his/her facial expression; (2) text-based analysis, by the detection of words and sentences typical of cyberbullying harrassements, oppression and stalking; (3) behavioural analysis by the detection of the keystroke dynamics and touch analytics. The statistical and generative models behind this tool are inspired by psychological models of criminal behaviour. To make them usable and testable in realistic scenarios, juridical and privacy implications are studied and a proposal to overcame them is also among the project results.


1) F. Balducci, D. Impedovo, N. Macchiarulo, G. Pirlo, "Affective states recognition through touch dynamics", in Multimedia Tools and Applications (2020). https://doi.org/10.1007/s11042-020-09146-4
2) S. Concas, S.M. La Cava, G. Orrù, C. Cuccu, J. Gao, X. Feng, G.L. Marcialis, F. Roli, Analysis of score-level fusion rules for deepfake detection, Applied Sciences, MDPI, 12 (15), 7365, DOI: 10.3390/app12157365, 2022
3) G. Orrù, D. Ghiani, M. Pintor, G.L. Marcialis, F. Roli, Detecting Anomalies from Video-Sequences: a Novel Descriptor, IEEE/IAPR 25th Int. Conf. on Pattern Recognition (ICPR 2021), Milano (Italy), 10-15th, Jan., 2021, https://arxiv.org/abs/2010.06407, DOI: 10.1109/ICPR48806.2021.9412855
4) Sansone, C., Sperlí, G. (2021). A Survey About the Cyberbullying Problem on Social Media by Using Machine Learning Approaches. In: , et al. Pattern Recognition. ICPR International Workshops and Challenges. ICPR 2021. Lecture Notes in Computer Science(), vol 12667. Springer, Cham. https://doi.org/10.1007/978-3-030-68787-8_48
5) G. Terrone, A. Gori, E. Topino, A. Musetti, A. Scarinci, C. Guccione, V. Caretti, The Link between Attachment and Gambling/Internet Addiction in Adolescence: A Multiple Mediation Analysis with Developmental Perspective, Theory of Mind (Friend) and Adaptive Response, Journal Personalized Medicine, vol. 11, no. 3, 2021; https://doi.org/10.3390/jpm11030228




Public Exposure


International Research Centre
on Artificial Intelligence (IRCAI)
under the auspices of UNESCO 

Jožef Stefan Institute
Jamova cesta 39
SI-1000 Ljubljana



The designations employed and the presentation of material throughout this website do not imply the expression of any opinion whatsoever on the part of UNESCO concerning the legal status of any country, territory, city or area of its authorities, or concerning the delimitation of its frontiers or boundaries.

Design by Ana Fabjan