- From: Chris Bizer <chris@bizer.de>
- Date: Wed, 17 Mar 2004 10:13:35 +0100
- To: "Jeremy Carroll" <jjc@hplb.hpl.hp.com>, "Patrick Stickler" <patrick.stickler@nokia.com>
- Cc: <phayes@ihmc.us>, <www-archive@w3.org>
Hi Jeremy, > > Note my section 4 is there mainly for Pat to trash and start over. > > I tend to think that an approach which permits you to trust (yes/no) a > number of named graphs of your choice gets close to the continuum approach > that is clearly a more accurate model of how people operate. e.g. as my > level of paranoia goes up I will trust fewer graphs. This is then reflected > in a classic logic with just t/f > No, I didn't mean that more paranoid people just trust fewer graphs, this is still the easy case :-) I meant, that an agent trust graph A more than graph B, e.g. he descides to trust graph A with a trust value of 0.8 on some trust scale and graph B with a value of 0.4. I think, we should at least teoretically be open to this kind of stuff. Chris
Received on Wednesday, 17 March 2004 05:12:13 UTC