Exploratory Analysis of Crime Behavior in the City of Medellin

Public Safety is one of the main fields of research in Smart Cities; this is because the sense of security in people can affect the development of a city or region. Crime is one of the main factors affecting security. Different hypotheses have been proposed to explain the behavior of criminal activities and formulate strategies that allow security forces to improve patrol plans and respond quickly to situations that affect the safety of citizens. Some hypotheses suggest that weather conditions influence people's behavior, while others point to a direct relationship between the unemployment rate and the increase in criminal activities. In addition, some theories allude that the poor infrastructure maintenance conditions of the place can incite crime. This work explores various hypotheses of criminal behavior in Medellin, one of the main cities in Colombia, and is recognized for having appreciable rates associated with crime. Databases are used with information on crimes in the city between 2015 and 2019, information from various meteorological stations located in different parts of the city, and a demographic variable obtained from the national statistical database. The analysis shows that about 30% of crime is concentrated downtown; the rush hour, which is between 6:00 p.m. and 8:00 p.m., is the time of day with the highest number of incidents; events such as salary payments and festivals also affect the metrics; moreover, an influence close to 35% of the unemployment rate is perceived in the crime variation. Finally, there is no evidence of a direct relationship between crime and temperature variation.