Estimating note phrase aesthetic similarity using feature-based taxicab geometry

Musicology is a growing focus in computer science. The comparison of musical phrases and the assessment of their similarity is a basic skill for musicians but is a rarely explored subject in the context of computer science. Computer science has a large variety of measures for the task of assessing similarity of different data types but very few can be applied when assessing music. In this article, we present a novel approach of assessment, specifically designed for measuring the similarity of the aesthetic of note phrases based on features extracted using jSymbolic applied in the context of taxicab geometry. In this study, dissimilarity or distance is proposed to be a function of several factors including note occurrences, chromaticism, note motion, among other factors. This study's results may be used in the development of aesthetic-based algorithmic musical composition.