Opportunities and challenges in rhythm research by using data-driven and culture specific approaches
Within the field of Music Information Retrieval a major goal is to automatically annotate music data, typically audio. But given that each musical culture thinks and expresses itself in a particular way we have to pay a special attention to this cultural specificity when annotating a particular music fragment. Most MIR research has focused on the western commercial music of the past few decades and the issue of multiculturality has not been a major concern. As soon as we open up our research target to the world’s music the cultural concern becomes fundamental if we want to develop IT systems that can be meaningful to particular cultural contexts. When we study specific music traditions, different from the western commercial music, we discover research opportunities and challenges that had not been identified before within the MIR community. Rhythm is a musical facet that is cultural specific and its computational study benefits from this approach. In the CompMusic project (http://compmusic.upf.edu) we are studying the Hindustani (North India), Carnatic (South India) and Turkish makam (Turkey) music traditions, and we are just starting to work on the Andalusi (Maghreb) and Beijing Opera (China) traditions. Each of these traditions offers great opportunities and challenges for the computational study of musical rhythm and in this talk I will present the current rhythm research being carried out in CompMusic related to these traditions and I will discuss our future research plans.