Hiya Dave... lookin for a new place to hang m'keyboard if ya'll will have me. I see yerself, Feyd and even David (Ctabuk) running around..can't be all bad. Ol WebNutz can be fun to play with... ran into each other here and there....
Anyways, the latter patent I mentioned is something Bill sent me over the weekend and I have just finished my first draft of an article about it. It is quite related to the earlier PaIR stuff and a recent MS publication as well
Using Search results
The crux of them is trying to teach the system to better understand natural language phrasings and relationships of words.
IMO it is simply more layers into the indexation/ranking process. You can see the problems it will start to have for the Black Hatters as far as using Markov type applications not to mention any movement away from a reliance on backlinks makes their life harder.
It offers search engineers a double benefit of tighter results and fighting the web spam at the same time.
Imagine a world where link spam, kw stuffing and kw density all went south?? Weeeeeeeeee .. fun stuff.
Strangely the patent I am looking at as well as the MS one, are somewhat using 'user sessions' as part of the evaluation process.... so how far off is 'Query Spam'? bwaaaa ha ha ha ha ha.....