Towards automatic assessment of musical knowledge

Ricardo Conejo, University of Malaga, Beatriz Barros, University of Malaga,

Automatic assessment of actual musical knowledge is not an easy task. There are different aspects including just cognitive knowledge, like recognizing the notes on a pentagram, and others that also require knowledge acquired with our senses, for instance recognizing the sequence and tempo of a set of notes played. We have developed the SIETTE authoring tool, representing a set of template assessment items based on a mashup of modified existing music technologies: (1) The Lilypond score language; (2) A javascript sound synthesizer; and (3) A javascript piano interface. Combining these technologies it is possible, for instance, to construct a question that produces the sound corresponding to a sequence of notes, and asks the user to recognize the notes and click on the corresponding piano keys as a response.