W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > February 2003

Re: Issues danc-01 Re: 2 formalities in RDF concepts

From: pat hayes <phayes@ai.uwf.edu>
Date: Mon, 3 Feb 2003 16:13:42 -0600
Message-Id: <p05111b1cba6494f2147a@[10.0.100.86]>
To: Dan Connolly <connolly@w3.org>
Cc: w3c-rdfcore-wg@w3.org

>On Tue, 2003-01-28 at 12:07, Dan Connolly wrote:
>>  On Tue, 2003-01-28 at 11:22, Jeremy Carroll wrote:
>>  > Hi Dan,
>>  >
>>  > before the WG discusses this issue I wanted to understand your concern.
>>  >
>>  > I believe that you think the concept described in
>>  >
>>  > 6.3 Graph Equality
>>  > http://www.w3.org/TR/rdf-concepts/#section-graph-equality
>>  >
>>  > is one that is relevant to RDF and should be described in RDF 
>>Concepts, but
>>  > that you take issue with it being labelled "Equality". Earlier 
>>drafts called
>>  > this concept "Isomorphism"
>>
>>  Well, that's a different way to resolve this matter, but
>>  the semantics doc doesn't use that notion of graph;
>>  it uses the notion of graph where the n-triples
>>  documents below definitely correspond to the
>>  identically same graph, even if the sets
>>  of triples don't share a blank node.
>>  So going that way would involve a change to
>>  the semantics doc that might be quite significant.
>
>After discussion with JanG, I think perhaps the change
>to the semantics doc might not be so significant.
>
>Just strike this bit:
>
>[[
>Graphs with isomorphic pictures in this sense are considered to be
>identical; this means that we do not distinguish sets of triples which
>differ only in the identity of their blank nodes. This slight abuse of
>terminology allows us to simplify the presentation by ignoring questions
>of re-naming of bound variables.
>]]
>
>since, in fact, in the definition of merge, we do in fact treat
>re-naming of bound variables.

Peter was right. I can make this change if y'all prefer but it will 
have other consequences throughout the document where I use this 
assumed identity. I will have to re-state and re-prove several of the 
lemmas. I honestly think this would be a mistake. Still, if it helps 
to keep things straight then I am willing to do this. Expect it to 
add about a week to the time to a final version of the semantics doc.

BTW, if we make this change then we will effectively have restored 
the 'problem' which gave us such trouble at the California F2F about 
18 months ago, and which we solved by adopting the graph syntax as 
definitive. In effect, we will have tossed away the graph syntax and 
replaced it with the N-triples syntax, thereby re-introducing bound 
variables and the need to refer to re-naming operations on merging. 
We then ought to go back through all the documents and purge all 
references to blank nodes as having no names, since they will have 
names.

There are really two ways to go. We can say that RDF syntax (I'll 
avoid the G-word for now) uses identifiers for blank nodes somehow; 
or not, ie blank nodes are blank. If we go the first way, then the 
concept of 'same graph' is effectively useless, since a graph with a 
systematic change to its bound variables is semantically 
indistinguishable from the original; and we have to talk everywhere 
about equivalence classes of graphs, in effect, or isomorphism 
between graphs. Also there are issues about scopes of these bound 
variables, which we put to bed 18 months ago and which it would be a 
chore, to put it mildly, to re-open at this stage. We decided in 
Sebastopol to go the second route, ie to identify graphs with 
indistinguishable blank nodes and to therefore not bother with 
notions of renaming, bound variable and so on.

My own preference is that we retain the advantages of this second way 
of doing things and re-write the concepts document wording to reflect 
the decisions we made in Sebastopol and which we have been using as a 
basis for our discussions ever since.

>Also, add graph to the list of terms imported from concepts.
>[[
>We use the following terminology defined there: uriref, literal, plain
>literal, typed literal, blank node and triple.
>]]

I am willing to do this IF the concepts doc is rewritten to say 
clearly and unambiguously that an RDF graph *is* a set of RDF 
triples, rather than any kind of 'graph' in any sense from 
mathematical graph theory. Otherwise I would prefer that the 
definition in the concepts document be replaced with a reference to 
the semantics document.

>I do think the term "Graph Equality" is misleading; I'd prefer
>Graph Equivalence for the RDF Concepts heading.

I disagree. If we stick to our guns here, we really do mean graph 
equality. It is up to us to define 'graph' so that equality comes out 
right, and we know how to do that.

We have had this worked out now for a very long time. Why screw it up 
at this stage?

Pat


-- 
---------------------------------------------------------------------
IHMC					(850)434 8903 or (650)494 3973   home
40 South Alcaniz St.			(850)202 4416   office
Pensacola              			(850)202 4440   fax
FL 32501           				(850)291 0667    cell
phayes@ai.uwf.edu	          http://www.coginst.uwf.edu/~phayes
s.pam@ai.uwf.edu   for spam
Received on Monday, 3 February 2003 17:12:05 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:55:47 EDT