Data analysis of european union states: youth behavior in digital world

Abstract

The paper deals with data analysis of young people behaviour (in age from 16 to 29 years) in digital world living in the European Union States. For this analysis, ten selected indicators from the area of digital world (focused for example on working with Internet, on social networks, on the Internet banking, on searching of information about travelling, about goods and services, e-mail communication and calls) and two economic indicators (gross domestic product per capita and unemployment rate) were chosen. Derived attributes were calculated in the data pre-processing phase. Selected algorithms of the agglomerative hierarchical clustering (as are the nearest neighbour method, the furthest neighbour method, the centroid clustering, the median clustering and the Ward method etc.) were used to find groups of similar objects (individual states of the European Union) based on the chosen indicators. Values of average coefficients of growth were used for clustering. The best results of clustering were achieved by the Ward‘s method; the data was divided into three clusters. Identified groups of European Union States were described by mentioned indicators.