Discovering spatio-temporal relationships in the distribution of building fires

This paper focuses on a systematic risk mapping for mitigation purposes, which plays a crucial role in the context of the emergency management. It applies principles of the knowledge discovery and data mining to support development of the fire risk model at a city level. The study offers a set of advanced methods of the spatial and spatio-temporal analysis that share the same goal - to unveil causal relationships in the incident data. Each of the methods, however, reveals different aspects of the relations, which represents a valuable source of information. The results of this research thus enhance understanding of the phenomenon being studied and enable more accurate risk maps to be created. (C) 2013 Elsevier Ltd. All rights reserved.