AnalyzingThoughtStorms (ThoughtStorms)


Wikis and HyperTexts should have interesting network properties. So what if you try to apply SocialNetworkAnalysis techniques to them? Not so much to the social but to the pages.

Conveniently ThoughtStorms (being UseMod) and other wikis spit out a TouchGraphWikiBrowser format file. Which we can use to get the structure. (More on this format Meatball:LinkDatabase)

So I knocked up a quick PythonLanguage /Script to pull it in and count the number of inlinks to each page.

Here's the link to see a list of ThoughtStorms pages sorted by number of inlinks (A kind of primitive PageRank) :

You can point the script at any other wiki that spits out TGWB format too. Use CGI parameters like those in the examples below. (NB : change to wikan2 for a very baddly drawn frequency graph.)

Examples :

** see also : BillSeitz:WikiNameNetwork


But these have a link-base which are wrapped in anchors so screw up at the moment, will fix.

I wonder what else we can look at.

And here is ThoughtStorms indexed together with it's environment - that is you can search all the pages that are linked to from ThoughtStorms (together with the pages in the wiki):

-- ZbigniewLukasiak

Wow! That's actually quite spooky. It's sort of the Kirlian photography of search isn't it? You get to see the aura around ThoughtStorms when you search for a term. I think I'm going to add a search box to it here. :-)

-- PhilJones

There is one undocumented feature - '!' in entry fields works as negation. So for example to see where 'PhilJones' is mentioned on the pages you link to you can search like this:

-- ZbigniewLukasiak

See also :