The Kiki-Bouba Challenge: Algorithmic Composition for Content-based MIR Research and Development

We propose the “Kiki-Bouba Challenge” (KBC) for the research and development of content-based music information retrieval (MIR) systems. This challenge is unencumbered by several problems typically encountered in MIR research: insufficient data, restrictive copyrights, imperfect ground truth, a lack of specific criteria for classes (e.g., genre), a lack of explicit problem definition, and irreproducibility. KBC provides a limitless amount of free data, a perfect ground truth, and well-specifiable and meaningful characteristics defining each class. These ideal conditions are made possible by open source algorithmic composition — a hitherto under-exploited resource for MIR.