Comments on: The future of social networking and graphs http://anders.tyckr.com/2008/01/17/the-future-of-social-networking-and-graphs/ Words of a startup CEO Fri, 21 Nov 2008 02:56:43 +0000 http://wordpress.org/?v=2.5 By: anders http://anders.tyckr.com/2008/01/17/the-future-of-social-networking-and-graphs/#comment-13 anders Wed, 06 Feb 2008 08:26:40 +0000 http://anders.tyckr.com/2008/01/17/the-future-of-social-networking-and-graphs/#comment-13 Thanks, good points. But I disagree that <em>my</em> music graph just has me in it, the way I see it is that my graph <em>is</em> the all of the peers of neighbors with similar taste. The difference between my and someone else in the same graph, is that it is centered around me. This is a real fuzzy concept, but I think some interesting thoughts arise when trying to apply this to different things that do not yet have a goto service like Last.fm has for music. Thanks, good points. But I disagree that my music graph just has me in it, the way I see it is that my graph is the all of the peers of neighbors with similar taste. The difference between my and someone else in the same graph, is that it is centered around me.

This is a real fuzzy concept, but I think some interesting thoughts arise when trying to apply this to different things that do not yet have a goto service like Last.fm has for music.

]]> By: liesen http://anders.tyckr.com/2008/01/17/the-future-of-social-networking-and-graphs/#comment-12 liesen Wed, 06 Feb 2008 01:11:48 +0000 http://anders.tyckr.com/2008/01/17/the-future-of-social-networking-and-graphs/#comment-12 For uses of music data; take a look at APML (www.apml.org) and http://tastebroker.org/. A site that extracts and uses attention data is http://idiomag.com -- plug in your last.fm username and you've got a customized magazine about your favourite artists. And regarding <em>your</em> music graph--which is alone not so useful because it's just you in it--, you really want your peers' nodes (the people with similar taste as you) in it to take advantage of recommender systems and the like. A question arises: what data should be extractable from a service? In this case (with music), there aren't a lot of harmful data that can escape a service, but there might be other cases in which the definition of harmful data is not so clear (e-mail addresses, etc. (do OpenID)). For uses of music data; take a look at APML (www.apml.org) and http://tastebroker.org/. A site that extracts and uses attention data is http://idiomag.com — plug in your last.fm username and you’ve got a customized magazine about your favourite artists.

And regarding your music graph–which is alone not so useful because it’s just you in it–, you really want your peers’ nodes (the people with similar taste as you) in it to take advantage of recommender systems and the like.

A question arises: what data should be extractable from a service? In this case (with music), there aren’t a lot of harmful data that can escape a service, but there might be other cases in which the definition of harmful data is not so clear (e-mail addresses, etc. (do OpenID)).

]]>