Re: computer-go: Human move ID process survey...

["GL7: David Elsdon" <gl7@xxxxxx>]

At 27/05/03 16:00, you wrote:
>  I'm doing some research into applying AI to Go and my thoughts have 
> turned to how humans play and emulating them rather than how to make a 
> computer play and then working from there.
This is what I did with my program GL7 - currently languishing at the 
bottom of the 9x9 ladder. I do a few board transformations to try and 
capture those elusive intuitions like:

    *  A measure of the influence of each stone on the board
    * Strings of stones and their liberties
    * Groups of stones
    * Boundaries between stones of the opposing players
    * Live and dead stones

then I run through a set of rules - the first rule that succeeds generates 
the move. Very simple !! No searching and no evaluation. Perhaps a bit too 
simple to be honest, but I think it's an approach worth pursuing. I have 
temporarily lost enthusiasm, but when I get going again ...

Cheers

David (UK 2kyu)