"similar pages" in google : how does it work?

Hi guys,

When I first type related:http://www.cataloguecity.co.uk in google, I get a short (but good :-)) list of websites all related to it.... fine .

But when I type related:http://www.catalogcity.fr, I get a larger list, but none of the links are actually related to this website at all !!

Does someone know how this feature works ?

Carote

Try it here:
http://www.touchgraph.com/TGGoogleBrowser.html

Applauses please!

Really, what do you think about this tool?

well...doesn't give me anything when I query my website (address above)!!! Empty page

This is very cool. I do not really understand it but it is cool. I am going through the instructions now. I do not get anything when I click on a node though.

This is very cool. I do not really understand it but it is cool. I am going through the instructions now. I do not get anything when I click on a node though.

Carote - This site requires the Java plug-in. Even then, it takes a while to first load up.

What I really don't understand is the signifigance of the Related: query.

How are these results determined?

What should it mean to me?

-Nate

Nice javascript. You get my applause all the way. This is my new favorite tool of the month! I have begun to understand the relationships, but still working it out.
Is it possible to have the program query internal links, outgoing links, and Google backlinks and make relationships between the website or sites?

Thanks for the link!

Ben

No problem loading on this end. I had to download the Javascript Run Enviroment thingy for it too load up. But its worth it, and if you are on anything slower than 56K it might be a problem.
I run a 256K speed and no issues and delays to bad.

Nice one here goes my applause to you.

There is a similar one at http://www.kartoo.com . That one is more visual I have to say. It needs flash plugin installed though.

They both look cool tools BUT what about the original question? Can anyone help? I would like to know too, how 'related' is worked out by Google. It has been bugging me for a while that I don't know.
Any help???

I poked around a bit, and all I can find are various people's theories.

I think it's a bit of "Secret Sauce"... My big question now is if it's important enough to worry about.

Some have postulated that it is a function of PR, others say editorial, still more claim both.

My personal theory is that it goes something like this: find pages which share a similar and high number (percentage?) of single words AND PR is similar. Much like what we understand the ranking alg to be...

Proof? We don't need no stinkin' proof.

-Nate

WEll, I don't know about keywords. One of my sites Gemfields.com has itself as well as ttcms portal, the portal software it is built with as similar. The only similarities to ttcms are really just the general layout and that we link to ttcms. So an editor would not have decided the similarity and PR and keywords are also different altogether. The other example is a freebi site i run ozfreebies.com and there the similarities are more freebie sites and then a GIFT site (no free keywords) an ABOUT search page for homepages, more freebie sites and then a Hotel and other unrelated sites again.
Just seems strange, so I just wondered. Does it make any difference to searches etc Prbably not but I am still curious. Where is this G G when we need him 8)