Geospatial visualization of indicators for the dynamics of innovation in an educational institution applying clustering techniques

Abstract

The activities of science, technology, and innovation are related to the execution of actions involving research, experimental development, support for education and training, provision of scientific and technological services, administration, and other management activities. In this context, the SENNOVA Research System of Servicio Nacional de Aprendizaje of Colombia dedicates human and technological resources to contribute to the country’s economic and social growth, looking to answer the need to develop Colombia’s productive sector. In turn, these contributions also generate social dynamics in which the activity at the institutional level can be represented as a kind of the complex systems studied by nonlinear Physics. These complex dynamics are suitable for visualization from the stochastic processes that lead to statistical distributions typical of complex systems. A data analytics model for the measurement and visualization of innovation indicators is being developed in the Antioquia regional branch of Servicio Nacional de Aprendizaje, where the distinct categories of science, technology, and innovation activities are graphically identified to facilitate the analysis of the results obtained from both descriptive statistics and data science. From the perspective of complex systems for representing these institutional social dynamics, clustering processes with techniques such as K-means grouping were implemented. Potential distributions determined by the conglomerates of management processes and productivity of the projects executed in the analyzed institution over a given period are identified as innovation indicators and subsequently classified using principal component analysis. Python Folium was used as a visualization tool to graphically generate comparisons between the different Servicio Nacional de Aprendizaje centers of the Antioquia regional branch in each period. The results show greater ease of interpretation and analysis of statistical results and data analytics in measuring indicators of science, technology, and innovation activities through the techniques employed in comparison with traditional data visualization tools.