https://www.cqu.edu.au/cquninews/stories/research-category/2015/using-software-to-predict-our-preference-for-a-song-or-melody

It would be handy to be able to predict an individual's preference for a song or melody in order to create personalised advertising or to pitch a new song to a prospective buyer.
That's according to CQUniversity Sydney IT lecturer Dr Rohan de Silva who recently filed a provisional patent at IP Australia following the successful completion of the honours thesis of his Bachelor of Engineering student Dean Strike.
The provisional patent ‘A Method and System to Predict the Preference of a Given Song or a Melody by a Person’, co-invented by Rohan and Dean, was based on an idea and concept provided by Rohan to the School of Engineering and Technology honours thesis project pool in 2014. CQUniversity is now searching for partners to commercialise this invention.
"This invention can be used in personalised adverting and marketing where a song track or a melody that a person prefers can be played in the advertisement," Rohan explains.
"This will be possible with Internet-based advertisements and advertisements sent to mobile phones or other handheld devices.
"The invention can also be used to deliver a new song to a prospective buyer who would very likely appreciate the song by a particular composer or a production company.
"As a commercial product, the invention will appeal to organisations that focus on cross-demographic profiling to achieve a sustainable competitive advantage."