Computational methods to study piano music in education context

In the last decades, music students, teachers, and institutions are embracing online education in an accelerated way. In this context, this master thesis aims to carry out a computational exploratory analysis about Piano Music Education’s new technologies from symbolic data. We propose technological approaches to various aspects of Piano Music Education in its most global view, from technique to accompaniment and composition. First, we suggest two automatic piano fingering systems. Second, we depart from one previous generative fingering system to explore the technical difficulty in the new Mikrokosmos dataset. Third, we model the relationship between a soloist and piano accompaniment suggesting descriptors in three dimensions: melodic, rhythmic, and harmonic. Lastly, we confirm that automatic composition systems can be biased to generate an individualised repertoire based on difficulty level. The presented work makes a step toward expanding current Music Information Retrieval approaches for Piano Music Education and opening new horizons and challenges for further research