Just wondering, does the semantic web encompass similarity searches on non-structured content such as music , video and images? For instance one blob is sufficiently similar to another blob to consider them equivalent, or to put it in more mundane examples:

1- musical piece number 1 is similar to musical piece number 2 except that one of them contains more improvisations

2- the symptoms look sufficiently similar to those belonging to disease ABC

A lot has been said about web 2.0 and, like most all-encompassing visions, ultimately more has been said than done. It seems that the semantic web will go the same way.