Mining Chordal Semantics in a Non-Tagged Music Industry Database

It is recognized that non-temporal MPEG-7 based MIR systems are both state-ofthe-art and diametrically unconducive to pitch analysis. Music patterns contain particular scales and chords that affect human emotion. Recognition of chordal patterns has a direct empirical association with tension in a song and how pretty, melodic or dark the song may be. Finding the scale and root of a musical piece is the prerequisite to finding chords and patterns and up until the authors’ recent research, scale and root mining was not possible. Until scale and key patterns can be found it is impossible to mine the economic viability of a song. This paper continues constructing an automatic decision support system that upon mining the correct scale and key of an untagged song in a database can now recognize the correct chordal progressions inherent in the song. Herein we present a system that can mine chordal progressions in an untagged music database.