Vadim - 1:42 am on Apr 14, 2005 (gmt 0)

data mining can often do a better job of deciding what is relevant than humans can.

It sounds like “Your Big Brother Data Mining Robot knows better than you what you need”. For this to be true, the data mining robot should have the life experience and knowledge of a particular human or at least a human expert in a particular subject. At present state of the art it is simply impossible because it means to solve the problem of the artificial intellect.

The keyword here is life experience, i.e. *history*.

So it seems that the Goggle patent is about the history and its importance in the search results.

This is the right direction but at present state of the art, it seems the only way to achieve the stability is to feed the data-mining robot with highly relevant criteria from the start. I.e. ask the human experts first.

The present Google instabilities seems show that they spend not enough efforts to select, or rather compose, really relevant criteria before enter them into their data mining robot.

Vadim.