Playing around with PeepAgg for a while, considering the concept behind “people aggregation”, than jumping in my mind to the sense or non-sense of digg, I come up with the idea of creating a kind of social network filter mechanism, which could solve all my problems I have with the web in the moment.

What’s your problem with the web pixelsebi?

Good question, thx for that. My problem is all that content out there, which is not usefull to me, but shows up at google, technorati, digg, flickr, del.icio.us - whereever I throw an eye on, to grab some information I need quickly.

I think the key issue is, that crawlers, pingservice driven search engines, user-driven-news sites and many more, don’t solve my problem at all: finding content, which really fits to my individual requirements.

In real life we have a proofed concept to solve problems like this: we call a friend of us, from whom we know, that he’s an expert on the topic we are investigating in and we get a good answer, because we know this guy and his strengths.

Allright, I’ve understood your problem - but where is the problem? So just visit the del.icio.us page from your friend and go ahead. Everthing is community driven now, let’s rock!

Yes - that’s kind a practicable way to get this problem reduced. But this doesn’t does it the way it should be in my eyes. The community diggs stories - allright - but who the f*ck is this community? These all are people I don’t know, I don’t share expierences with these folks and I realy absoluty can not assess them, to get an idea of the quality of their recommended content.

But I think, excactly this is important to get what you want in nowadays of the web. You have to bring in your human advantage in having the ability to assess people you know in seconds and get conclusions from that, in regard to what you are looking for.

Seems to me, you allready have a idea of a solution for that?

Remeber XFN? Most of us, have it implemented in our blog software (without knowing and using it?) XFN is a microformat created to describe my social network, when I create links to people I know in any way.

And that’s the point, where it gets started. First of all, we need a XFN crawler, which fits our needs in creating a mapping of our social network - indepent from where we have this information. Important is only the thing, that we describe it with XFN.

So the first step of a application I have in mind is, to setup the sources of XFN marked content, which describe my social network. From this point, I need a crawler how creates a kind of map of my social network, I can work with. I wanna set emphases on parameters, which are important for my investigation (friend or not, e.g.) and create a ad-hoc filter, which I would trust.

After I have build this mapped social network, I try to enrich the information about it, by having a synchronisation with possible hCard resources, to get as much information as possible for every single node of my social network. You know, that XFN describes only the relation between these nodes. The data about the people has to be fetched elsewhere (even if the XFN could be part of an hCard)

After this sync is done, my filter is ready to be applied on a service, which is driven by people. Let me take for example digg. If digg would have an open API, I now would use my filter to do one of the following operations:

• show me only stories, where parts of my social network are envolved in (added or digged the story)
• show me every storie, which has at least one digg, from a friend of my first level social network
• show me every story, which was digged by e.g. James or his co-workers (who will think similar to him)

… and so on. I could now do several kind of operations on services like, digg or del.icio.us or even technorati, by defining how close the relation has to be with my social network. My knowledge about and experience with the people is the key in this mechanism to find content, which will solve my problem.

And it is even better, than our real life model. When I know, that for example James is working in a company, which is operating in the topic I am investigating in, I can add every co-worker of James to my request, because the chance is high, that they together have the best sources on that topic.

And I think, this would only be the beginning of possible szenarios. Visualisation of this network, with all my request and results I have done - growing relations of my network, which possible proof old results and so on, would be very exciting!

So what do you think of this approach? Possible? Sensefull? Allready there?
Send me feedback, write in the comments or trackback to me, if you continue this idea on your own blog.