Eurostat is proud to announce the winners of the classification challenge competition of the European Statistics Awards Programme, which focused on developing an automated solution to classify occupations described in online job advertisements.
The European Statistics Awards Programme seeks streamlined and efficient solutions to support Eurostat’s work on web intelligence for statistical purposes. This challenge invited participants to develop new methods that automatically assign class labels (based on a set taxonomy) to online job advertisements within a dataset that Eurostat provided.
The competition dataset included 26 000 anonymised online job ads collected from approximately 400 websites across the EU. These ads were gathered and anonymised by the Eurostat Web Intelligence Hub, which has collected and classified over 200 million job ads since July 2018.
The challenge attracted 121 teams comprising 228 individuals from 37 countries.
We are happy to reveal the winners in all 3 award categories:
Accuracy Award is given to the top 3 teams with the highest scores for best methods in correctly assigning class labels to job ads from a given dataset. The first-prize winner: TNO_AI_LAB.
Reusability Award is given to the top 3 teams with the highest scores for the most thoroughly described and documented solutions, with the most scalable, well described and open approach. The first-prize winner: TheClassifiers.
Innovativity Award is given to the top 3 teams with the highest scores for most innovative approach. The first-prize winner: TheClassifiers.
Take a look at the infographic to see who won second and third place.
The winning solutions, some of which incorporated artificial intelligence, are presented on the competition website.
Congratulations to all the winners!
