jeremy_goodrich - 9:01 pm on Jun 17, 2003 (gmt 0)

Microsoft has published some incredible research in collaboration with a university in Beijing, China.

The interesting bit is that they were looking at other types of correlations than the ones that the other players look at to map 'relevance'.

Another interesting bit is that Microsoft is one of the first large US corporations to incoporate fuzzy systems into their products, and that China has the world's largest number of fuzzy scientists.

Google's technology is based on a probabilistic model which is a subset of fuzzy logic, as determined by 'probability & the hypercube' by a professor named Bart Kosko.

In my mind, better math = better engineering. If I were Google, I would seriously rethink my bias against non traditional systems & fuzzy logic, and toss the binary thinking that may constrain their ability to return optimal web results.