Data science techniques applied to analysis of incidents registered by the 1-1-2 Canary Islands emergency services
Date:
July 08-11, 2018. Co-authored with Pérez-González C, Roda-García JL, Rosa-Remedios C and González-Dos-Santos T. Technical Program, page 273.
Abstract
The study of alerts received in the emergency services is a very important issue in order to know different aspects of the time and spatialdistribution of alerts in a given region. In particular, the Emergency and Security Coordinating Center (CECOES) 1-1-2 of the Government ofthe Canary Islands records detailed information about the incidents that are reported by the citizens through phone calls. Due to the high volume of information generated over the time in this process, it is needed to apply big data techniques to obtain statistical measures and results ofinterest. We have used the statistical software R and different libraries (Shiny, Highcharts, Highmaps) to present the data information in different interactive dashboards (time series charts to analyze the timeevolution, tree classification of the sanitary emergencies, geospatial representations of incidents density distribution, etc.) and to propose several predictive and classification models. In this work, we illustrate some illustrative and valuable results in studying the incidents in theregion during the last years.