Result Prediction for the Eurovision Song Contest

Thesis Type Master
Thesis Status
Currently running
Student Adrian Johannes Marxer
Thesis Supervisor
Research Field

In this work, the result (ranking) for songs performed at the Eurovision Song Contest (ESC) should be predicted. While typical hit song prediction is done on music charts such as the Billboard Top 100, this thesis should be done with restriction to the ESC. Therefore, methods used for hit song prediction on charts should be applied and if necessary adapted to the ESC. Audio features should be combined with other features, such as the singing language and the public appearance of the artists (e.g., social media, news, ...). In other words, this thesis combines internal and external features to predict a given ranking of songs.