Analítica descriptiva para la comprensión del estado actual del fenómeno delincuencial en el Ecuador

Abstract

This study carries out a practical application of data analysis techniques to generate intelligence that supports informed decision-making on public safety and crime prevention in Ecuador. The stated objectives were: a) To conduct a theoretical study on descriptive analytics and its related concepts, b) To analyze the INEC database related to crime in Ecuador, c) To develop descriptive analytics models that allow the analysis of the crime phenomenon in Ecuador. The study applies the agile MAMBO methodology for the analytical processing of a dataset on arrests and police apprehensions in Ecuador. It uses 79,609 records with 23 fields covering demographic, geographic, and types of crime information. The inductive approach enables insights generation from the data to support decision-making in citizen security. Techniques include descriptive statistics, data visualization, ARIMA time series modeling, and aggregation algorithms. Data cleaning, integration, and transformation are applied; then, dashboards with indicators about the demographic profile of offenders, geographic comparisons, temporal evolution, and patterns by hours are generated. The results reveal a predominance of young repeat male offenders. Guayas, Pichincha, and Manabí have the highest criminal activity, identifying crime peaks during nighttime. A trend forecasting model is also constructed, suggesting stability with slight fluctuations. Thus, the study demonstrates the value of Big Data analytics in comprehensively characterizing the crime phenomenon Keywords: descriptive analytics, crime, MAMBO.