Geospatial Distribution of the Quality of the Coffee of Minas Gerais by Methods of Data Mining Techniques

Abstract

The state of Minas Gerais is located in one of the most privileged Brazilian regions for the production of high-quality coffees. Due to its large territorial extension and environmental variation, it favors the production of specialty coffees with a great diversity of flavor and aroma. These differences are related to the peculiar characteristics of each microregion, particulary climate variations, altitude, latitude and production systems. Although the relationship between environmental characteristics and coffee productivity in different biomes is already well explored, its relationship with beverage quality still requires research. Given the need for more information on areas with potential for special coffee production, this paper proposes to associate data concerning to sensorial quality of coffees in the Cafés of Minas Quality Contest, with climatic and geographical variations of each state region, applying association rules for data mining, used to understand the aptitude of the state through the generation of rules that describe the most relevant patterns in the data. The climatic, geographical and classificatory information of the coffee samples is the result of the database of the contest participants in 2007 and 2008. The results demonstrate that with data mining techniques it was possible to associate the regions with the greatest potential with the climatic and geographical characteristics of the environment.