Trust accounting: a possible approach

Hello!

I have very-very preliminary version of a paper on trust accounting. There are 
two basic assumptions in it:
- trust is measured in (o;c) pairs, where o is our opinion on the subject, 
while c is certainty. The main operation is averaging (weighted summ).
- there is some good-enough domain and name system that expresses some 
relations between message emitters and groupings of them (.com-like DNS zones 
aren't good enough, because they are flat, e.g. www.samsung.com and 
www.samsung-monitor.com are formally unrelated; old-fashioned .edu zone seems 
to be good enough, but it is a relict of Young Universe)

Because of the second assumption the paper will have mostly theoretic value 
today. Anyway, discussion on the list seems to raise very similar issues, so 
I'm looking for co-authors here. This pie is too big.

http://blogs.plotinka.ru/gritzko/accounting.ps.gz

Thank you.

--

	Viktor S. Grishchenko
	research assistant
	Institute of Physics and Applied Mathematics
	Ural State University
	Yekaterinburg, Russia