Title of article :
Collaboration Perspectives for Folk Song Research and Music Information Retrieval: The Indispensable Role of Computational Musicology
Author/Authors :
Peter van Kranenburg، نويسنده , , Jorg Garbers1، نويسنده , , Anja Volk، نويسنده , , Frans Wiering، نويسنده , , Louis P. Grijp، نويسنده , , Remco C. Veltkamp، نويسنده ,
Issue Information :
روزنامه با شماره پیاپی سال 2010
Abstract :
Background in Folk Song Research (FSR). In the past century melodic variation caused by oral transmission has been studied within the discipline of Folk Song Research (FSR). Also, various systems have been developed to categorize large collections of folk songs. Since the 1940s many attempts have been made to design automatic systems to categorize melodies. However, after several decades, no strong theories of oral transmission, and no generally applicable classification systems have yet emerged. Currently, many cultural heritage institutions give high priority to the digitization and unlocking of their (musical) collections. Background in Music Information Retrieval (MIR) and Computational Musicology (CM). In the field of Music Information Retrieval (MIR) methods are designed to provide access to large bodies of music, while in the field of Computational Musicology (CM) computational methods are designed to study musicological questions. These developments stimulate a new interest in the questions of FSR. However, very few CM and MIR studies take the particularities of orally transmitted melodies into account. Aims. Better collaboration between MIR, CM and FSR (or Musicology in general) will enrich Musicology with new methods to study existing problems and MIR with better understanding of music. Main contribution. By surveying relevant achievements of the disciplines, we show a gap between MIR and FSR. To bridge that gap we provide promising directions for research based on current developments, as well as a collaboration model in which CM serves as an intermediate between FSR and MIR. Implications. MIR should go beyond provided ‘ground truth’ data in implementing and testing models that generated those ground truths. The concepts used in folk song research, ‘tune family’ in particular, should be modelled, providing MIR a musically informed implementable model and FSR an enriched understanding of those concepts.
Keywords :
Computational Musicology , Music information retrieval , Tune Family , Folk Song Research , Tune Classification , Tune Identification
Journal title :
Journal of Interdisciplinary Music Studies (JIMS)
Journal title :
Journal of Interdisciplinary Music Studies (JIMS)