Google’s secret lab
The rumors on Google using human human operators to tweak their algo seem to be true:
It’s one of the best kept secrets of Google. It’s a mystery on Webmasterworld. Also in Europe (France) they don’t know what to expect from that odd URL http://eval.google.com. Click it and you get …nothing. The site reveals itself only if you have the proper login and if you use a network known by Google. Residues of Eval.google are found on the web, but the full content of the mystery site has never been published before.
Credits for the link go to Razvan Pop
IN YOUR FACE YAHOO!
I wouldn’t jump on that
ahahaha really nothing to worry about for Yahoo…
Razvan Pop is trying to point out that this leaked secret is in in fact a part of a bigger PR campaign with the role of getting image for Google. I think that indeed there is nothing to worry for Yahoo
They’ll have plenty to worry once Google will become once again undisputed leader in serving relevant results.
Undisputed leader in serving irelevant rumors. Check your spell check
Just because a RATer Hub was unveiled doesn’t change a thing.
I see that as a space ship powered with human slaves instead of fuel and I told u that
Aha… and I see the SEO bussiness going down the drain, since humans won’t be susceptible to the usual SEO tricks, I see Google finaly realizing that the relevance problem is one without a solution (algoritmicaly speaking) and this is a huge leap ahead for them… at this point they’re way ahead in the game, so… vroom, vroom Google, bye bye … the rest of wannabie search engines.
What was that about the spell check? here’s one “irrelevant”
U might have a point mate. Any algo can be at least partially deducted through euristhics and that’s where the humans might help…but after all you talked bout a learning algo.
Anyway too tired to explain you now that it’s quite impossible to find a statistically relevant set of RATerS for every field of knowledge.
Ah yes… the (in)famous learning algo… It’s quite easy actualy, all you need is something along this line: http://www.paulgraham.com/better.html .
In the case of the Bayesian filter users tell the filter what is/isn’t spam, this is a simple binary problem, relevance is a bit more complicated but it could be done. And an even more complicated algo should be able to extrapolate from a few (limited number) of domains…
©2005 Google – Searching 8,058,044,651 web pages. yes, billion.
this number will tripple by 2010
I cant even comprehend the amount of manpower it would take to check the relevency of this many sites.. Can you?
The spiders are here to stay. Algorithyms are a nessesity.