摘要：A new online archive of encoded fado transcriptions is presented. This dataset is relevant as the first step towards a cultural heritage archive and as source material for the study of songs associated with fado practice using empirical, analytical and systematic methodologies (namely MIR techniques). It is also relevant as a source for artistic purposes, namely the creation of new songs. We detail the constitution of this symbolic music corpus and present how we conceived of and implemented a methodology for testing its internal consistency using a supervised classification system.
关键词：symbolic corpus;methodology;music information retrieval
Loading...
