DataFest Germany 2026 at the University of Mannheim
DataFest Germany 2026 brings students from across Europe to Mannheim
From March 6–8, 2026, the University of Mannheim hosted the 9th edition of DataFest Germany, a data science competition that brought together Bachelor’s and Master’s students from universities across Germany and beyond. The event was jointly organized by the School of Social Sciences and the Mannheim Center for Data Science (MCDS) in cooperation with partners from Ludwig-Maximilians-Universität München and BERD@NFDI.
At the beginning of the competition, 11 teams with members from more than 20 universities received a previously unseen dataset provided by Wikimedia, containing detailed information on the 10,000 most visited Wikipedia pages in 2025 across the ten largest language editions, including page metadata, page views, and editing activity. Over the course of nearly three days, participants explored the dataset, developed models and visualizations, and worked intensively to uncover meaningful insights.
After two days of coding, analysis, and discussion, the teams presented their findings in a four-minute “All-or-Nothing” final presentation, moderated by Prof. Florian Keusch. The jury – Jan Karem Höhne (University of Hannover & DZHW), Polina Galkin (SAP), Fabian Block (Nesto Software), and Johannes Blumenberg (GESIS) – faced the difficult task of selecting the winning teams.
The awards went to:
- Best Insight:Duisburg Analytica (Ruhr University Bochum, University of Marburg, University of Bamberg), whose cluster analysis of different Wikipedia editor types impressed the jury.
- Best Use of Outside Data:Noverfit (Linköping University, TU Dresden, Leipzig University, RPTU, DBU), who combined Wikipedia metadata with Google BigQuery data and information on German parliamentary seats to examine Wikipedia as a political news source.
- Best Visualization:Chai Alliance (Hertie School), whose presentation analyzed editing behavior across languages and topics and explored whether an open encyclopedia can remain neutral in a conflicted world.
In parallel to the student competition, doctoral and postdoctoral researchers from the KODAQS Academy participated in a separate challenge focusing on data quality insights. KODAQS Team 1 won the Best Data Quality Insights award after comparing more than 36,000 cross-language Wikipedia articles and demonstrating that many differences across language editions stem from editorial decisions rather than structural metadata.
Throughout the weekend, participants were supported by academic mentors and professional consultants. A Career Fair provided opportunities to connect with potential employers, and the organizers ensured that participants were well supplied with food and refreshments during the intensive competition.
DataFest Germany was organized by Dr. Florian Bemmann, Prof. Florian Keusch, and a dedicated team of researchers, staff members, and student volunteers who contributed to making the event a success.
The organizers would like to thank the event’s sponsors Nesto Software GmbH, Lidl Stiftung & Co. KG, Atruvia AG, and dm drogerie markt GmbH, as well as all volunteers for their generous support.
The next DataFest Germany will take place in Munich in 2027.
