QuickTopic free message boards logo
Skip to Messages

TOPIC:

Amazon networks

(not accepting new messages)
6
Steve YostPerson was signed in when posted
01-14-2003
04:48 AM ET (US)
We'd love to hear about it!
5
Kevin Marks
01-14-2003
04:43 AM ET (US)
I plotted the Blog ecosystem and got a convincing power-law distribution that saturated early (understandably, as they won't find all the 2 and 1 incoming links blogs with their spider)
I got similar results form Movie box-office data - I keep meaning to write this up.
4
Steve YostPerson was signed in when posted
01-08-2003
12:57 PM ET (US)
Steve, that's a very good point. Given that, it seems the map might contain some false positives, i.e. implied clustering of tastes that doesn't exist. OTOH I don't think it refutes the dichotomy that's seen in Valdis' map.

Adina, other maps such as plain link connectivity have been done (URL anyone?), but I think it'd be hard to to blogroll mapping in practice without a standard for blogrolls, i.e. it'd be hard to parse them automatically. But it'd be great, wouldn't it? It might be a good discovery mechanism too. I often feel too limited in my blogroll cluster, wondering if I'm missing another nice cluster just over the hill.

Valdis responded in email after he backtracked from a referrer. I think he's OK with me posting this because he was looking for the discuss link. Here's what he said:

Actually this book network does NOT show a power law distribution...

And a link was drawn when either book of a buddy pair listed the other as a buddy. Go to "American Jihad" and you will see "What Went Wrong" listed. Good observation... thanks for the feedback!
Edited 01-08-2003 12:59 PM
3
Adina Levin
01-08-2003
09:36 AM ET (US)
It would be fascinating to map blogroll links. That would reveal some very interesting cluster patterns, I think.

I've mentioned the idea to Valdis but haven't taken initiative to do anything about it yet.
2
steve himmer
01-07-2003
06:16 PM ET (US)
One thing that isn't revealed by a map like this is actual usage, and interpretation. For instance (as I wrote in David Weinberger's comments), Samuel Huntington's Clash of Civilizations is a book I've read more than once, but entirely disagreed with. I've used both as a student and a teacher with others who disagree with the books conclusions. But that won't show up in a map, only in person-to-person, detailed research. It's one thing to know what books people read, but knowing what ideas characterize 'the left' or 'the right' in relation to those books is something else.
1
Steve YostPerson was signed in when posted
01-07-2003
05:16 PM ET (US)
Blog post: http://www.quicktopic.com/blog/archives/000192.html

What do you think?
Edited 01-07-2003 05:16 PM