Classification and spatial analysis of fires in the operational dispatch management system

Abstract

The article explores the potential of applying modern information technologies, specifically big data processing methods, artificial intelligence, and machine learning, to analyze operational information accumulated within the Operational Dispatch Management System (ODMS) of the Main Directorate of the State Emergency Service of Ukraine in Lviv Region. Operational since 2007, this system has amassed a substantial volume of data on emergency incidents, encompassing details about response forces and resources, event types, personnel, technical equipment, and more. The study aimed to develop an information model for the analysis and prediction of risk scenarios. Several stages were implemented within the research framework: preliminary data structuring and cleaning, sample formation based on classifiers, detailed categorization of fire types, geocoding of addresses, and visualization of results through heatmaps. Additionally, an approach to semantic analysis of textual fields and event clustering by type was proposed, enabling flexible sample generation for subsequent investigations