Developing a data analytics platform to support decision making in emergency and security management

Published in Expert Systems with Applications, 2019

Recommended citation: Pérez-González CJ, Colebrook M, Roda-García JL, Rosa-Remedios CB. "Developing a data analytics platform to support decision making in emergency and security management". Expert Systems with Applications 120, 167-184 (2019) https://doi.org/10.1016/j.eswa.2018.11.023

Abstract

The Emergency and Security Coordinating Centre is the public service responsible for managing all the incidents registered in the Canary Islands. More than 7 million records have been collected in the last decade with more than twenty observed variables for each incident, which comprise more than 140 million data. All these data emanate from different islands and municipalities but with very marked differences. The study in this paper presents complete and novel research about geographical and temporal incident distribution, which may be of interest to emergency services managers and people responsible for designing public policies concerning security and health matters. We have developed an analytical web platform that features several dashboards with statistically significant results, by island, municipality, etc., and incorporating certain external data sources regarding social and economic issues which allow us to study the relationship between these factors and incident distribution at different geographical levels. Certain specific results are presented and illustrated in order to show all dimensions of data analytics that not only significantly improve companies’ and organizations’ processes, but also demonstrate how data analytics competency relates to decision making performance. Several statistical models that forecast and classify the incidents are proposed to illustrate the potential of statistical modelling in the study. The application has become into a crucial strategic tool for the organization because it helps in decision making processes. The information provided is highly modular and allows for the future inclusion of new features in order to provide larger and improved data analyses.