- From: John Black <JohnBlack@deltek.com>
- Date: Tue, 19 Oct 2004 10:15:35 -0400
- To: "John Black" <JohnBlack@deltek.com>, <Patrick.Stickler@nokia.com>, <otto@math.fu-berlin.de>
- Cc: <eric@w3.org>, <pfps@research.bell-labs.com>, <www-rdf-interest@w3.org>
Patrick Stickler wrote: >> It may be more useful to ask, what form of subgraph of a graph >> of RDF statements might constitute an optimal body of information >> about a resource, given a particular URI denoting that resource; >> or some such. In my previous post, this doesn't seem quite right either: > The problem I have with the phrase "optimal body of information" is > that it doesn't seem to include an objective against which the body > of information is optimized. Thus I think "minimal *denotative* > content" (or "minimal *denotative* body of information") is better > because it specializes optimal to minimal and provides an objective > function to be optimized, namely "denotative". In other words, it > must be both minimal *and* denotative. Without the objective function > in an optimization problem, the optimum is unconstrained, see > http://www.nist.gov/dads/HTML/optimization.html The goal is to optimize (minimize) size, while still denoting the resource. Say the URI is http://example/people/JohnBlack. What is the smallest graph that denotes me. If the graph just includes my weight of 190 pounds, it describes me and it may be minimal but it doesn't denote me. If it includes my entire autobiography it denotes me but it is not minimal. Just to say a graph is an "optimal body of information" seems to leave out the primary objective of denoting. John Black
Received on Tuesday, 19 October 2004 14:15:37 UTC