Neuro-Symbolic Approach for Tantrum Monitoring and Prevention in Individuals with Autism Spectrum Disorder: A Protocol for Virtual Agents

Autism Spectrum Disorder (ASD) poses unique challenges, with individuals often experiencing difficulties in emotional regulation, leading to disruptive behaviors such as tantrums. This research protocol outlines the development of an innovative system designed to monitor and prevent tantrums in individuals with ASD, employing a neuro-symbolic approach. The proposed system integrates neural networks and symbolic reasoning to improve understanding and prediction of tantrum episodes. Leveraging real-time physiological and behavioral data, collected through wearable devices and user input, the system employs machine learning algorithms to detect patterns indicative of imminent tantrum events. Furthermore, a symbolic reasoning system interprets these patterns, taking into account individualized factors. The outcomes of this study are anticipated to contribute valuable insights to the growing field of technology-assisted interventions for neurodevelopmental disorders.