Some holiday morning notelets ....

1. The social graph in action. I felt a tremor in the social graph this week. A bundle of my Facebook befrienders attended the CETIS conference. I was suddenly aware of status lines, notes, imported blog entries. I had a sense of some of what was discussed and could follow up if I wanted. It happened in the background. It was like the weather: I had a sense of what was happening without having to do much investigation. Incidentally, CETIS have done a nice job in collecting some of the network amplification of the conference on the website: blog posts, del.icio.us bookmarks, and so on.

2. The social graph, not. Facebook's flatness does not very well accommodate our layered and multidimensional social lives. A lot to talk about there, but this is still a holiday morning notelet .... To pick a simple and relatively straightforward example: what to do with an unwelcome invitation to be a 'friend' from your boss? I assume we will see a more nuanced way of managing the ways in which we present ourselves emerge over time. Which raises issues about how we port or share our represented identities, something that we do not do well now. The social graph is site-specific.

3. Net, web, graph. Tim Berners Lee gave the social graph expression a lift yesterday in a post about the evolution of our networked environment. He talks about a net/web/graph stack. The 'net' allowed us to address computers directly, abstracting away from the underlying connection paths. The 'web' allowed us to address documents, abstracting away from the machines on which they reside. In each case, new and unanticipated value was built on the navigable spaces the net and the web created. The 'graph', Tim Berners Lee suggests, allows us to work with the things that documents are about, friends, flights, proteins, customers and so on, abstracted away from the documents or sites themselves. If represented appropriately, and he uses the example of FOAF, applications can combine and recombine data about things across multiple documents and sites. So, an application could combine what various sites know about me and my relationships. So yes, in these terms, the social graph meets the semantic web. Of course, we have yet to see whether Facebook believes that the social graph is actually greater than the Facebook graph.

Comments: 1

Nov 22, 2007
alanp

I agree totally with your point on nuancing - until SocNets are a lot better at simulating our real life behaviour they are going to be vulnerable to "cheat" gaming.

Re Sir Tim....seems like the GGGG is a specialised case of the Semantic Web to me, ie a far smaller taxonomy set - just the social net data - is defined..