ABSTRACT
This chapter is intended to understand how moods might affect music composition by either allowing for or obstructing innovative use of pitch. It is part of an ongoing research project focused on developing artificial intelligence systems that can aid music composition in the creation of new mood-based music. While it is clear that the framework for pitch transitions in the song are related to the top codewords found from the happy dataset, this does not hinder the use of innovation. While the top codewords found in the middle of the graph show high levels of interconnectivity, there are around thirty instances of once-visited codewords which surround these commonalities. The availability of music-related big data has made supervised machine learning a popular method in current research. One area of interest is the machine learning algorithm's ability to predict at such a high level of accuracy given the likelihood that unwanted pitches permeate the datasets, as shown in the "Billie Jean" example.
