[rdfweb-dev] A definition of 'knows'

Julian Bond julian_bond at voidstar.com
Sun Aug 3 13:21:30 UTC 2003


Dan Brickley <danbri at w3.org> wrote:
>Also brings up the question (discussed in IRC recently) about whether we
>say that foaf:knows is symmetrical, ie.
>whether ?x foaf:knows ?y can be possible without ?y foaf:knows ?x.
>Intuitively I find this quite appealing, but it has sensitive corner
>cases (eg. alzheimers)...

LOL! I've got a note to write about this, re scutters. IMHO, foaf:knows 
is directional and also almost meaningless. That doesn't mean that it's 
useless. So A knows B plus B knows A is worth more than just A knows B. 
And "knows" means whatever the creator thinks it means but it does imply 
at least "some" relationship. So when it comes to readers they can't 
draw many more conclusions and in turn make up whatever it means for 
themselves. But they should take note of where a link is said to be 
symmetrical by the two participants.

I think the line from foaf:knows to rel:some_complex_relationship[1] is 
rather like a similar argument that comes up often about rating schemes. 
You can get lots of benefit from a binary rating scheme 0 or +1 without 
ever saying what +1 actually means. People repeatedly come back and say 
"why can't we have negatives", "I don't understand +1, why can't we have 
X, Y and Z on scales of -5 to +5", "I don't like X,Y,Z we should have 
A,B,C and 0-10". And so on. To which I say *pish* ;-) You've got 0 or 
+1, get over it.

[1] Just occurred to me. Could family relationships be defined by a 
series of vectors on a 2D graph of the family tree? uncle = 1,1 
grandfather = 0,1:0,1 cousin = 1,1:0,-1
-- 
Julian Bond Email&MSM: julian.bond at voidstar.com
Webmaster:              http://www.ecademy.com/
Personal WebLog:       http://www.voidstar.com/
M: +44 (0)77 5907 2173   T: +44 (0)192 0412 433



More information about the foaf-dev mailing list