Dataset Forró em Vinil

In the work described in both the paper <em>Measuring disruption in song similarity networks </em>and the doctoral thesis entitled <em>Caracterizando e Quantificando Inovação em Redes de Similaridade Musical </em>(only available in portuguese), Falcão et al. developed a study of disruption measurement over a collection of a Brazilian music tradition called Forró. This dataset contains all the audio information used during analysis. 25,605 audio files were used to build a song similarity network, from which disruption information was derived. <em><strong>list_of_songs.txt</strong>,</em> <strong><em>mfccs.csv</em> </strong>and <strong><em>latent_representations.csv</em></strong> contain an indexed list of all the audio files analysed, their MFCC-based feature vectors and a simplified latent representation of the MFCCs generated by an autoencoder developed during the study, respectively. Data from all these files can be mapped by using the indexes (i.e., the i-eth line in <em>mfccs.csv</em> refers to the feature vector for the i-eth song informed by <em>list_of_songs.txt</em>) <strong><em>similarity network - unweighted.gexf</em></strong> contains the similarity network built according to the similarities calculated using the feature vectors. The network is available in a GEXF (<em>Graph Exchange XML Format)</em> format, and can be visualized with softwares like Gephi.