Automatic Music Score Difficulty Classification.

Learning a musical instrument is challenging in many different aspects. One important aspect is the motivation of the music students. One common practice to keep music students engaged is to use songs that the students like. However, identifying an appropriate music score for the music students that is compatible with the specific skill level and music taste of each student can be a challenging task. Defining the difficulty of each score manually is a complex task, as it requires theoretical knowledge and mastery of a particular musical instrument and music theory. In this context, the automatic music score difficulty classification task can help music teachers and students by automatically classifying music scores according to their difficulty. The main contribution of this work is to evaluate the use of two approaches based on the musicXML files and different classification algorithms for this task concerning three different music instruments, namely the violin, the piano and the acoustic guitar. Our experiments show interesting results for all the musical instruments.