Voice of the Streets: a platform for urban violence detection based on social sensing

Urban violence poses a significant challenge for citizens and municipalities, necessitating strategies to effectively disseminate official information while residents seek timely access to critical information. This study presents ‘Voice of the Streets (VOS)’, an Information and Communications Technology (ICT) service designed to detect and classify violence in social media messages and understand its effects in physical environments. Using Design Science Research (DSR) as the methodology, this study developed a platform that integrates social sensing with a machine learning model to categorize the severity of violence. Results demonstrate the platform’s feasibility in integrating external data sources, processing multiple data instances, and supporting the visualization of violence patterns.