Euan at The Obvious sums up his view of the “blogroll or not blogroll” debate as “I just like following winding paths”.
At the same time Ton Zijlstra picks up on various experiments with Social Network Analysis of the blogosphere and draws the important distinction between a map based on “who knows you” - i.e. an analysis of inbound links (and I would say by implication trackbacks and comments) and a map of “who you know” based on outbound links.
This feels time to send a request to the Lazyweb - a graphical web-based tool that takes a URL and presents some kind of graphical depiction of the incoming and outgoing networks it discovers…
Updated Matt Jones writes Bridging The Bubbles about similar ideas applied to finding the bridiging points between clusters of particular political (or other) views… amongst the sites he references is Valdis Krebs’ analysis of booklinks on Amazon Divided we stand? : Political patterns on the WWW
Updated again Just found Reputation and Conversation in Blogging and Network Topology at TIG’s Corner [via Doc Searls]
Update 3 GoogleBrowser looks like it might be part of the way there in terms of the display…. [via Ross Mayfield]
Facebook
January 8th, 2003 at 11:06 pm
Bridges and Bubbles
Matt Jones has an entry today Bridging the bubbles that reminds us how we find new vistas, and adds some cautions about what can go wrong when we don’t. I’m piqued by
January 9th, 2003 at 4:43 am
I have to agree with Euan: GoogleBrowser is a nice pretty picture, but it doesn’t tell you anything about the quality of the links, and that actual quality is important. Before any such map makes any sense, we need to know if the site making the link is ‘reputable’ (in our subjective view) and whether their link is an accident, a reflex, or an actual endorsement, and if which, then how much or even what mix of all three! Basically, all you /can/ do is blindly explore
January 9th, 2003 at 4:51 am
http://www.kartoo.com/flash.php3?vis=1&langue=en&ca=&l=&m=&fd=&bo=&tr=&version=4 is yet another approach, but I find it a bit more useful because it will at least divulge the /reason/ for the link. It’s also flash-based which is a bit more performance on most machines than a java applet.
January 11th, 2003 at 10:10 pm
If you have a human panel manually assess links and subsets of a network, they will wade through the data, weighing dozens of factors to form their judgements. That’s the standard to which we should hold SNA tools.
September 25th, 2003 at 9:16 pm
valdis krebs needs a weblog…
valdis krebs’ inflow mapping knowledge creation weaving well-formed webs… haitech haiku™ ©2003 judith meskill This morning I received a comment on my Notable Judiths - Judith Donath post of 31 August 2003 from Valdis Krebs [whom I met virtu…