The prediction of the seasonal Influenza is important for Health Organizations in order to reduce the disease’s impact. By allocating resources and planning vaccination campaigns, the risk of infection can be decreased. However, collecting and processing data about flu cases takes up to two weeks which leads to increased forecast uncertainty. Twitter data is available in real­time and reflects the development of the flu, helping to improve the forecast. Additionally, the spatiotemporal dependencies in the data can be used to gain information about the disease’s spreading.

The team

  • Hendrik BerkemeyerHendrik Berkemeyer
  • Pascal NietersPascal Nieters
  • Hristofor LukanovHristofor Lukanov





  • Prof. Dr. Gordon Pipa
  • Prof. Dr. Kai-Uwe Kühnberger
  • PD Dr. Helmar Gust
  • Prof. Dr. med. Dipl.-Phys. Bertram Scheller


  • Kira Hildebrandt
  • Tobias Petri
  • Artur Czeszumski
  • Olga Vakhovska
  • Olena Forynna
  • David Hoffmann
  • Devrim Celik
  • Maria Sokotushchenko
  • Renato Garita
  • Hiroshi Sawada
  • Frank Mehne
  • Khanh Le
  • Turan Orujlu
  • Andrew Melnik