This release contains the scores of uzun hava in **kern format. The dataset is used in the masters thesis:
Şentürk, S. (2011). Computational modeling of improvisation in Turkish folk music using variable-length Markov models. Master’s thesis, Georgia Institute of Technology, Atlanta.
Please cite the publication above in any work using this dataset.
This version has 77 uzun havas from different regions of Anatolia, Iraq and Caucasus. It consists of 10849 notes in total, in 8 makams. For more information about the dataset, please refer to the thesis.