an Entity references as follows:
A classifier trying to capture the notion of structure in an audio piece. This classifier should be subsumed by more specific classifiers: speech/music segmentation, structural music segmmentation (intro, verse, chorus, etc.).