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

Re: williams-01, proposal to close (revised)

From: Graham Klyne <GK@NineByNine.org>
Date: Wed, 09 Apr 2003 16:23:07 +0100
Message-Id: <5.1.0.14.2.20030409154836.02dd25b0@127.0.0.1>
To: Jeremy Carroll <jjc@hpl.hp.com>, w3c-rdfcore-wg@w3.org

For my counter-proposal, skip to end of message.  For rationale, read on...

At 08:34 09/04/2003 +0300, Jeremy Carroll wrote:
>OLD: The underlying structure of any expression in RDF
>OLD: can be viewed as a  directed labelled graph, which consists of
>OLD: nodes and labelled directed arcs that link pairs of nodes
>OLD: (these notions are defined more formally in section 6). The RDF
>OLD: graph is a set of triples:
>
>NEW: The underlying structure of any expression in RDF
>NEW: is a collection of triples, each consisting of a subject, a
>NEW: property and an object. A set of such triples is called an
>NEW: RDF graph (defined more formally in section 6).
>NEW: The structure can be illustrated by a directed node-arc diagram
>NEW: in which each triple is represented as a node-arc-node link
>NEW: (hence the term "graph").
>I prefer NEW except for "directed" which should modify arc not node.
>Suggest:
>JJC: The structure can be illustrated by a  diagram with directed,
>JJC: labelled arcs linking nodes.

Hmmm... maybe.


>UNCHANGED:
>[[
>Each property arc represents a statement of a relationship between the things
>denoted by the nodes that it links, having three parts:
>
>a property that describes some relationship (also called a predicate),
>a value that is the subject of the statement, and
>a value that is the object of the statement.
>]]
>
>maybe that should have been changed.
>My understanding is that the resolution of the issue uses the terminology
>triple/node/URIref etc for the graph
>and
>statement/resource etc for the domain of discourse.

I think it's very clear from our discussions that "nodes" of a graph are 
its "subjects" and "objects";  cf. 
http://www.w3.org/TR/2003/WD-rdf-concepts-20030123/#section-rdf-graph
which I understand to reflect the graph discussion.  Otherwise, yes, I agree.


>Suggest:
>
>JJC: Each triple represents a statement of a relationship between the
>JJC: things denoted by the nodes that it links. Each triple has three parts:

That's better that mine.

>JJC: a subject node representing the resource that is the subject of the
>statement,
>JJC: a property URI reference that represents the relationship of the
>statement (also called a predicate),
>JJC: an object node that is the object of the statement.

Ummm... this appears to be inconsistent:
(a) use "subject" or "node", not both (see above).
(b) the subject node *represents* a "resource that is the subject", but an 
object node "is the object ..."

I think my wording here can be improved, but I don't think yours does 
it.  Following the lead of:
[[
6.1 RDF Triples

An RDF triple contains three components:

     * the subject, which is an RDF URI reference or a blank node
     * the predicate, which is an RDF URI reference
     * the object, which is an RDF URI reference, a literal or a blank node

An RDF triple is conventionally written in the order subject, predicate, 
object.

The predicate is also known as the property of the triple.
]]
-- http://www.w3.org/TR/2003/WD-rdf-concepts-20030123/#section-triples

Here's another attempt:

[[
Each triple represents a statement of a relationship between the
things denoted by the nodes that it links. Each triple has three parts:

a subject,
an object, and
a property (also called a predicate) that denotes some relationship between 
the subject and object.
]]



>UNHCANGED:
>[[
>The direction of the arc is significant: it always points toward the 
>object of
>a statement.
>]]
>again the statement word is misplaced.
>
>JJC (suggest):
>[[
>The direction of the arc is significant: it always points toward the 
>object of
>the triple.
>]]

Fair point.  Or just?:
[[
The direction of the arc is significant: it always points toward the object.
]]


>OLD: -
>NEW: The nodes of an RDF graph are its subjects and objects.
>suggest "subject nodes and object nodes"

That's not what your normative text says :^)  (see above)

>OLD: The meaning of an RDF graph is the conjunction (i.e. logical AND)
>OLD: of all the statements that it contains.
>
>NEW: The assertion of an RDF triple says that some relationship,
>NEW: indicated by the property, holds between the subject and
>NEW: object of the triple. The assertion of an RDF graph amounts
>NEW: to asserting all the triples in it, so the meaning of an RDF
>NEW: graph is the conjunction (logical AND) of all the
>NEW: statements it contains. A formal account of the meaning of RDF
>NEW: graphs is given in [RDF-SEMANTICS].
>
>Neither seem to obey the rules on triple vs statement.
>Also, I think it is better to avoid the assertion word where possible
>(justification: social meaning resolution) (i.e. I believe there is avoidable
>risk associated with adding new occurrences of the A-word).

I don't recognize any such risk associated with the word "assertion" -- 
it's simply something that is claimed or presumed to be true.  I think the 
risk arises only when we start to discuss how such claim is established.

>Suggest:
>
>JJC:
>[[
>The meaning of an RDF triple is the statement it represents.
>  The meaning of an RDF graph is the conjunction (logical AND) of
>the meaning of all the triples it contains.
>A formal account of the meaning of RDF graphs is
>given in [RDF-SEMANTICS].
>]]

I think this fails the intent of this (non-normative) section, which is to 
provide some lead-in/background to the concepts that underpin RDF.  "The 
meaning of an RDF triple is the statement it represents" feels rather 
circular to me.  I prefer the form of words I proposed (which IIRC were not 
mine, but offered by Pat).  But maybe one could be tighter about use of 
some terms:

[[
The assertion of an RDF triple says that some relationship,
indicated by the property, holds between the things denoted by
subject and object of the triple.  The assertion of an RDF graph
amounts to asserting all the triples in it, so the meaning of
an RDF graph is the conjunction (logical AND) of the statements
corresponding to all the triples it contains.  A formal account
of the meaning of RDF graphs is given in [RDF-SEMANTICS].
]]

>Overall for section 3.1 my prefered text is as follows:
>
>[[
>  3.1 Graph Data Model
>
>The underlying structure of any expression in RDF
>is a collection of triples, each consisting of a subject, a
>property and an object. A set of such triples is called an
>RDF graph (defined more formally in section 6).
>The structure can be illustrated by a  diagram with directed,
>labelled arcs linking nodes.
>in which each triple is represented as a node-arc-node link
>(hence the term "graph").
>
>*picture*
>
>Each triple represents a statement of a relationship between the things
>denoted by the nodes that it links. Each triple has three parts:
>
>+ a subject node representing the resource that is the subject of the
>statement,
>+ a property URI reference that represents the relationship of the statement
>(also called a predicate),
>+ an object node that is the object of the statement.
>
>The direction of the arc is significant: it always points toward the 
>object of
>the triple.
>
>The nodes of an RDF graph are its subject nodes and object nodes.
>
>The meaning of an RDF triple is the statement it represents.
>  The meaning of an RDF graph is the conjunction (logical AND) of
>the meaning of all the triples it contains.
>A formal account of the meaning of RDF graphs is
>given in [RDF-SEMANTICS].
>]]

My collected counter-proposal:

[[
3.1 Graph Data Model

The underlying structure of any expression in RDF
is a collection of triples, each consisting of a subject, a
property and an object. A set of such triples is called an
RDF graph (defined more formally in section 6).
This can be illustrated by a node and directed-arc diagram,
in which each triple is represented as a node-arc-node link
(hence the term "graph").

   *picture*

Each triple represents a statement of a relationship between the
things denoted by the nodes that it links. Each triple has three parts:

   a subject,
   an object, and
   a property (also called a predicate) that denotes a relationship.

The direction of the arc is significant: it always points toward the object.

The nodes of an RDF graph are its subjects and objects.

The assertion of an RDF triple says that some relationship,
indicated by the property, holds between the things denoted by
subject and object of the triple.  The assertion of an RDF graph
amounts to asserting all the triples in it, so the meaning of
an RDF graph is the conjunction (logical AND) of the statements
corresponding to all the triples it contains.  A formal account
of the meaning of RDF graphs is given in [RDF-SEMANTICS].
]]

#g


-------------------
Graham Klyne
<GK@NineByNine.org>
PGP: 0FAA 69FF C083 000B A2E9  A131 01B9 1C7A DBCA CB5E
Received on Wednesday, 9 April 2003 13:02:38 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:56:54 EDT