The PhD thesis "Biometrics for Safety and Health in Real-World Contexts" ex-plores the application of behavioral biometrics for monitoring and enhancing quality of life, with a particular focus on preventing bullying and cyberbullying among young individuals. The research leverages innovative technologies such as Human Activity Recognition (HAR) and continuous authentication through touch dynamics to construct behavioral models capable of timely identifying anomalies and potential threats to safety and psychological well-being. This study is part of the PRIN-2017 "BullyBuster" project, an interdisciplinary in-itiative aimed at combating bullying through artificial intelligence and computer vision technologies. During the research, an Android app and a web platform were developed, integrating a questionnaire designed to collect biometric and in-teraction data during use, allowing for the classification of users based on their responses and behavioral activities. Through data collection in school and university settings, datasets were created to train and validate machine learning models capable of identifying behaviors associated with bullying and victimization. The results demonstrate how behav-ioral biometrics can serve as a valuable tool for supporting health and safety in real-world environments, offering a prevention and monitoring system adaptable to contemporary needs.
La tesi di dottorato "Biometrics for Safety and Health in Real-World Contexts" esplora l'applicazione delle biometrie comportamentali per monitorare e migliora-re la qualità della vita, con particolare attenzione alla prevenzione del bullismo e del cyberbullismo tra i giovani. La ricerca sfrutta tecnologie innovative come il Human Activity Recognition (HAR) e l'autenticazione continua basata sulle di-namiche di tocco, al fine di costruire modelli comportamentali capaci di identifi-care tempestivamente anomalie e potenziali minacce alla sicurezza e al benessere psicologico. Questo studio si inserisce nel progetto PRIN-2017 "BullyBuster", un'iniziativa in-terdisciplinare volta a contrastare il bullismo attraverso l'intelligenza artificiale e tecnologie di computer vision. Durante la ricerca sono stati sviluppati un'app Android e una piattaforma web, integrando un questionario progettato per rac-cogliere dati biometrici e di interazione durante l'utilizzo. Questi dati hanno per-messo la classificazione degli utenti in base alle loro risposte e alle attività com-portamentali. Attraverso la raccolta di dati in contesti scolastici e universitari, sono stati creati dataset per l'addestramento e la validazione di modelli di machine learning, capa-ci di identificare comportamenti associati al bullismo e alla vittimizzazione. I ri-sultati dimostrano come le biometrie comportamentali possano rappresentare uno strumento prezioso per supportare la salute e la sicurezza in ambienti reali, offrendo un sistema di prevenzione e monitoraggio adattabile alle esigenze con-temporanee.
Biometria per la Sicurezza e la Salute in Contesti Reali
GATTULLI, Vincenzo
2025
Abstract
The PhD thesis "Biometrics for Safety and Health in Real-World Contexts" ex-plores the application of behavioral biometrics for monitoring and enhancing quality of life, with a particular focus on preventing bullying and cyberbullying among young individuals. The research leverages innovative technologies such as Human Activity Recognition (HAR) and continuous authentication through touch dynamics to construct behavioral models capable of timely identifying anomalies and potential threats to safety and psychological well-being. This study is part of the PRIN-2017 "BullyBuster" project, an interdisciplinary in-itiative aimed at combating bullying through artificial intelligence and computer vision technologies. During the research, an Android app and a web platform were developed, integrating a questionnaire designed to collect biometric and in-teraction data during use, allowing for the classification of users based on their responses and behavioral activities. Through data collection in school and university settings, datasets were created to train and validate machine learning models capable of identifying behaviors associated with bullying and victimization. The results demonstrate how behav-ioral biometrics can serve as a valuable tool for supporting health and safety in real-world environments, offering a prevention and monitoring system adaptable to contemporary needs.| File | Dimensione | Formato | |
|---|---|---|---|
|
Final_draft_thesis - Revisionata 23-05_VG-DI-APsigned (1).pdf
accesso aperto
Licenza:
Tutti i diritti riservati
Dimensione
6.56 MB
Formato
Adobe PDF
|
6.56 MB | Adobe PDF | Visualizza/Apri |
|
Final_draft_thesis - Revisionata 23-05_VG-DI-APsigned (1)_1.pdf
accesso aperto
Licenza:
Tutti i diritti riservati
Dimensione
6.56 MB
Formato
Adobe PDF
|
6.56 MB | Adobe PDF | Visualizza/Apri |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/355112
URN:NBN:IT:UNIBA-355112