Formalistic Modelling Based on Pattern Recognition Applied to the Knowledge and Human Talent Sector in Ecuador

The purpose of this study has been to analyze the data set about the education received by high school graduates in Ecuador. The outcome may allow the recognition of behavior patterns and the corresponding relationship between their employment and associated economic activities. Initially, we conducted a qualitative evaluation of the methodologies and the freely available data mining tools. Subsequently, the data analysis has been applied on the central repository of the Knowledge and Human Talent sector of the Ecuadorian governing entity. Several classification algorithms, as well as association techniques have been performed using the CRISP-DM data mining process with R-Studio. For the grouping, we used the K-means algorithm. For the association, the A-piori algorithm has been used to encounter potential occurrences in the data set. Then, Pentaho Data Integration platform and PostgreSQL have been applied, in order to implement the study. The results demonstrate the dynamic generation of the social composition, with some detected issues, such as unemployment, underemployment, informal occupation, and even a surplus of professionals in certain areas, giving a conclusion of the relationship with their economic activity. Finally, the obtained information may allow the development of public policies that could improve the Knowledge management, as well as the productive matrix of the country.