Most people have many more than one page online that "represents" them, even if it's just a Facebook page and a MySpace page. This has been addressed quite nicely with XFN's
rel="me", especially now that there are tools like the Social Graph API to process this data efficiently.
This redundency phenomenon isn't limited to people, though. It seems to be a fact of life that anything in the real world is going to have more than one page about it on the web. Take events, for example: Hayseed Dixie will be playing my local venue the Colchester Arts Centre on July 22nd. There's a page about this on the venue's own site, plus a page on Eventful, a page on Last.fm and many more. The band itself has an official website and a MySpace page as well as profile pages on numerous music websites. The venue too has its own website along with profile pages on all of the event listing sites.
This suggests that the clustering of several URLs that represent the same thing is more generic a problem than that of personal relationships. This is reinforced by the fact that the social graph API handles "me" links in a special way, distinct from the other XFN edges. Would it be valid to use
rel="me" to link together the multitude of pages about an event, a band, a venue or other constructs like companies, clubs and schools? Although the other XFN edges aren't necessarily applicable, the the Social Graph API would support querying this with no changes. While it'd probably take a while for two-way links to emerge, it wouldn't take much work for sites like Last.fm and Eventful to include rel="me" on their links to bands' official websites and MySpace pages. Eventful already has links of various sorts about bands, venues and events which could be annotated in this way.
What do you think? Is this overloading the semantics of XFN's
me relationship, or a logical extension of it? More importantly, is this even useful?