Labour market attractiveness in the EU
In this work we present a framework developed for the European Big Data Hackathon 2017. This framework is divided in two parts: an exploration part, which is aimed to better understand the EU global labour market and to capture its heterogeneity; and an inferential part, whose goal is to establish associations between characteristics of the EU labour market and indicators designed to capture important aspects of the labour market (i.e. Skills mismatch, Mobility and Emigration). For the exploration part, we developed the concept of Labour Market Attractiveness, which consists of a combination of socio-economic and demographic variables from Eurostat datasets. Using data mining techniques, such as social networks and clustering analysis, we showed that this combined set consistently captured the country-level heterogeneity in the EU, forming well-defined clusters. For the inferential part, we used model selection analysis and weighted network correlation analyses to establish associations between the characteristics of the EU labour market and the labour market indicators. We argue that the combination of both exploration and inferential parts can disentangle the complex dynamics of the EU labour market and help setting effective policies to tackle typical problems of a fast-changing global labour market environment. [+]
