Keeping fans and the media updated on the Bundesliga’s star players is a key strategy in helping to build and improve the league’s brand and reach. Humans and machines are now working together closely to further improve player profiles: Every player in the Bundesliga will have an online profile on the league’s website and app, with biographical facts such as their nationality, birthday and age as well as information on their career and most recent matches. These machine-generated profiles are created by a Natural Language Processing (NLP) engine and are updated automatically after every matchday to provide the latest details about a player’s performance.
Due to the extent of standardised information, artificial intelligence (AI) can carry out this task far more quickly than any human editor could manage, given the sheer volume of data to be processed: 18 teams, each with a matchday squad of 18 players, for a total of 324 profiles. Nevertheless, cooperation between humans and machines is vital to ensure that the whole process runs smoothly.
The project is a three-way collaborative venture between Sportec Solutions (STS) – itself a joint venture of the DFL Deutsche Fußball Liga and Deltatre – which provides the data feeds, Retresco, the agency responsible for the NLP engine, and DFL Digital Sports.
As a system integrator, Sportec Solutions has expanded and prepared various data sources and contents and, with the interpretation specifications developed, laid the foundation for the “football intelligence” of the NLP engine. Each profile will initially be available in German and English, with Spanish set to follow in the future. Of course, technology never stands still, so the DFL is constantly working on upgrades and improvements. Further iterations will optimise the use case and develop additional applications. The player pages have already been launched online, with the app to follow. Further expansion is planned, with profiles in the pipeline for every Bundesliga 2 player.