One Nice Data Science Story

The first big data hackathon organized by the European Comission and EUROSTAT (statistical office of the European Union), took place between March 13th and 16h, 2017. in Bruxelles, Belgium. The event gathered teams from all over Europe to compete for the best data product combining official statistics and big data to support policy makers in one pressing policy question facing Europe. Teams of three members were proposed by the National Statistical Institutes and included members of the national data science community, staff of the statistical office or a partnership of both. There were 22 teams competing from 21 EU member countries.
There were five objectives: #1 Solve statistical problems by leveraging algorithms and available data; #2 Identify best of class European data scientists; #3 Promote and accelerate big data for statistics initiatives in Europe; #4 Promote partnerships with research community and private sector; #5 To produce innovative products and tools, including visualisations.
The policy domain selected was skills, education and lifelong learning which has an important role in the 10 priorities of the European Commission. The exact policy question for which the teams had to support the policy makers and analysts was the following: "How would you support the design of policies for reducing mismatch between jobs and skills at regional level in the EU through the use of data? Demonstrate your work through a tool that performs data analytics on the provided data and potentially additional data sources." The applications had to necessarily use a big data source and should ideally also have used European official statistics. Eurostat provided access to 5 datasets which the teams used to develop their prototypes, ranging from millions of web scraped jobs ads and CVs as well as several official statistical surveys. There were two evaluation panels: statistical panel combined from EU institutions experts and academia, and industry panel combined from representatives of technical vendors.  The winning solution, by the Croatian team consisting of Leo Mršić, Robert Kopal and Igor Kaluđer, was named "EU Skills & Jobs Explorer". It is a web application consisting of 12 visual, interactive dashboards enabling the policy analyst to gain insight into 4 key areas: skill supply, skill demand, skill gap and policy actions to close the gap. While many teams opted to use one of technical stacks provided by industry sponsors, Croatian team based their product on Google Cloud Platform on the backend side and Tableau for the frontend side.
Winning prototype can be accessed at live at: URL: eubdhack.in2data.eu; USER: [email protected]; PASS: <due to privacy policies can be sent only via email request>