This software adds an interesting take on personalized surfing – the USER has control.
Will be interesting to see if GYM buys this tech and incorporates it as a labs beta
http://www.web-ic.com/
http://www.expressnews.ualberta.ca/article.cfm?id=7507 |
Doesn't this have to do with the difference between treating each search query as independant (mutually exclusive) or as being related?
If you take Google's results currently, searches are completely independant. Searching for something one after the other has no impact on either SERP. However, in the day and age of machine learning and AI in general it is possible to discern patterns from user behaviour on a much more miniscule level.
This basically means that as a user refines his/her search query, the system reacts to the differences between each query. |
Have you heard of Pico (http://www.blinkx.com/press/2006/20060307.html) or Watson (http://www.intellext.com/) ? |
This is in a sense learning from the PROCLIVITIES of the searcher.
This tech is client-side, so in theory it is not like a persistant cookie or a sign-in database or users searchers that some search engines are keeping. |