String Methods for Folk Tune Genre Classification

In folk song research, string methods have been widely used to retrieve highly similar tunes or to perform tune family classification. In this study, we investigate how various string methods perform on a fundamentally different classification task, which is to classify folk tunes into genres, the genres being the dance types of the tunes. A new data set Dance-9 is therefore introduced. The different string method classification accuracies are compared with each other and also with n-gram models and global feature models which have been proven to be useful in previous folk song research. They are shown to yield similar results to the global feature models, but are outperformed by the n-gram models.