I am working on a school project in which I would like to use gracenote's mood classifications to recomend music which will help a user to exercise more often.
In order to achieve this, we would like to be able to interpret the moods in terms of how they relate to one another. Some googling around suggests that the 100 level 2 moods correspond to a 10x10 grid with the x axis ranging from calm to energetic and the y axis ranging from dark to positive.
We would like to be able to quickly relate a mood's text identifier or id to it's location in this grid so that we can treat the moods as a 2D spectrum rather than completely independent values. I have found a grid describing the mood taxonomy on page 18 of the presentation found here, http://www.slideshare.net/swissnexSF/peter-dimaria-and-chingwei-chen-largescale-machine-listening-and-automatic-mood-labeling-for-music-discovery-in-consumer-applications.
My question is if there is an easy way to convert the 2 level mood ID to its coordinates in this grid. If I could access a list of mood ID's formatted in this grid, that would be a great help, or if the mood's grid location is somehow encoded into it's ID, that would be cool. Otherwise, I will probably have to use the image linked above and manually create a dictionary mapping mood texts to their coordinates.
Thanks for any help!