In other articles I’ve mentioned audio indexing of most frequently used words in podcasts and videos, and now I am saying this sexy growing hashtag applies to music tracks now playing, as an example, someone plays a song in their cloud player, the lyrics are already parse and indexed for the song, so the Amazon cloud player already can provide the lyrics that are most used, and produce a cloud of tags, more used are bigger. Then overlapping that they are able to give the index across all songs now playing, so now we have a dynamic tag cloud with words from lyrics getting bigger, and smaller as they are played in their cloud player.
The user is also able to click a word and drill down to get a list of all the songs currently playing with the lyric word. One caveat, some songs group words together, so it might not make sense out of context. Amazon could apply some rules in the AI engine that specify if certain lyrics are grouped together, then they constitute one tag.
This feature would be another great way to diacover new songs for the music aficionado.
An extension of this is to get the music sound foot print, sound waves from tracks, and overlay with songs now playing, so if you like the beat or rhythm of a song, you can listen to other songs with similar rhythm or beats now playing, another way to find new songs. The matching rhythm can be an automatic shuffle filtered by genre, if specified.