Real-time emergency event detection system for public safety using multi-source data

Public safety is an essential service offered in smart city projects to provide better safety and security for individuals and city infrastructure. The advancement in the field of Information Technology and the Internet of Things created much scope for using smart applications in the city to enhance the quality of service, leading to a better life in cities. This digitization generates a large amount of data within the city from distinct sources like social media, IoT, sensors, any user-generated content from smart applications. The data generated within the city are analyzed to discover valuable insights for producing better data-driven decisions and predictions, that are more crucial for efficient city administration. Making quick decisions and early predictions of crimes by real-time analysis of data help the smart policing system to provide better services in the city. This paper describes the scope of real-time big data analytics for finding appro-priate predictions and making quick decisions for public safety. A real-time big data analytics framework using multiple data sources is proposed for the smart policing service in the smart city environment. The framework is used to design a real-time emergency events detection system to help city administrators in taking quick actions for the safety of people and city infrastructure. The proposed system achieved an average accuracy of 73% for emergency event classification.