Re: rdf in atom
Story Henry
There are two problems though:
(1) Rather than having an IETF document explain how to make explicit
a mapping of every instance of the atom format to rdf it would be
better to create an ontology for atom and specify an official GRDDL
transform for it.
(2) I think the semantics of atom are wrong in the draft. It would
have helped somewhat to consult on the work of the atom-owl [1] group
before putting this spec together.
The key problem is the value of the subject:
[[
<atom:entry>
<atom:id>http://example.com/a</atom:id>
<atom:title>Test</atom:title>
</atom:entry>
and the map above as a child of at:entrymap, the following triple
would be implied;
<http://example.com/a> <http://purl.org/dc/elements/1.1/title>
"Test" .
]]
By making the value of the subject be the resource identified by the
id, you are going against what at least two
independently built ontologies for atom consider to be the value of
the subject of the entry.
The Atom-Owl ontology [2] has a XQuery and XSLT2 transform the above
which renders that as
[] a awol:Entry;
awol:id "http://example.com/a"^^xsd:anyUri;
awol:title "test" .
awol:id is a functional property, but it is not equivalent to
owl:sameAs . [3] This is important because it is what allows atom to
be metadata describing different states of a resource over time and to
be merged into one coherent database. Since atom allows two entries
with the same id to be contained in the same feed, for example
<atom:feed>
...
<atom:entry>
<atom:id>http://example.com/a</atom:id>
<atom:title>Gold increases</atom:title>
<atom:updated>2008-06-13T18:30:02Z</updated>
<atom:content>The price of gold has just gone up</atom:content>
</atom:entry>
<atom:entry>
<atom:id>http://example.com/a</atom:id>
<atom:title>Gold value rises</atom:title>
<updated>2008-06-14T02:30:02Z</updated>
<atom:content>The price of gold has just gone up by 20%</
atom:content>
</atom:entry>
...
</feed>
Your mapping would create a merged rdf graph described in N3 as
[] a awol:Feed;
awol:entry <http://example.com/a> .
<http://example.com/a> dc:title "Gold increases", "Gold value rises" .
awol:content "The price of gold has just gone up by 20%",
"The price of gold has just gone up" ;
awol:updated "2008-06-13T18:30:02Z"^^xsd:dateTime,
"2008-06-14T02:30:02Z"^^xsd:dateTime .
At this point it will be impossible in the rdf graph to know:
- which title goes with which content
- which title goes with which time stamp
- which time stamp goes with which content
yet that information is very clear in the atom document. It cannot be
therefore that you have correctly captured the semantics of the atom
format.
I would therefore urge you rather to help out on the atom owl work.
That has stalled for a while now, either because it is very stable or
because people have not yet pushed it far enough to demand more. In
any case it just requires a little push to take it to the next level,
which would be to make it the official ontology of atom.
Henry
[1] http://groups.google.com/group/atom-owl
[2] https://sommer.dev.java.net/atom/
[3] I think it is eminently discussable wether awol:id should be an
owl:DatatypeProperty or an owl:ObjectProperty . It used to be a
DatatypeProperty, but Elias Torres argued for
a DatatypeProperty. I would welcome more discussion on this issue.
On 1 Jul 2008, at 08:00, Dave Pawson wrote:
>
> http://xml.coverpages.org/draft-nottingham-atomtriples-00.txt
>
>
> AtomTriples: Embedding RDF Statements in Atom
> Mark Nottingham and Dave Beckett (eds), IETF Internet Draft
>
> A version -00 Internet Draft for "AtomTriples: Embedding RDF
> Statements
> in Atom" has been published through the IETF process. specification
> describes AtomTriples, a set of Atom (RFC 4287) extension elements for
> embedding RDF statements in Atom documents (both element and feed),
> as well as declaring how they can be derived from existing content.
>
>
> regards
>
> --
> Dave Pawson
> XSLT XSL-FO FAQ.
> http://www.dpawson.co.uk
Story Henry
On 2 Jul 2008, at 01:53, Mark Nottingham wrote:
> On 01/07/2008, at 5:58 PM, Story Henry wrote:
>
>> Nice work. I welcome the effort to find a mapping from atom to rdf.
>>
>> There are two problems though:
>>
>> (1) Rather than having an IETF document explain how to make
>> explicit a mapping of every instance of the atom format to rdf it
>> would be better to create an ontology for atom and specify an
>> official GRDDL transform for it.
>
> rather a lightweight way to move statements around in Atom.
>
> IMO, the problem with trying to define or use an Atom ontology has
> always been that most authors and extension writers don't have RDF
> in mind; while extracting semantics from the world around you is
> interesting, there also needs to be a way to shove them around with
> demonstrable intent, so that there isn't any ambiguity about what's
> happening.
If you are looking for a lack of ambiguity, then why not endorse the
process of creating an official semantics for atom, such as atom-owl ?
The advantage of this is that it would bring all the work from the
Semantic Web side to bear on atom. A GRDDL transform would then create
a standard, and unambiguous mapping from atom to RDF. This would allow
discussion of extensions to make a lot more sense.
Then an extension proposals such as this one will be clear and
unambiguous, because it will be clear what is being spoken about, and
it will be clear what the subject of the rdf extension is about, and
how the extension relates to the rest of the atom spec.
> WRT using GRDDL or other out-of-band mechanisms -- the overhead of
> processing GRDDL is IMO not acceptable in some cases,
The GRDDL need not be the only way people extract meaning from the
atom format. Specialised atom rdf parsers could I imagine come along
and avoid the transform step. The GRDDL would be useful as a way to
test the ontology.
> and furthermore, it's again not explicit/self-describing.
? Is what you are doing self descibing? The RDF would is a lot more
self describing than the atom xml... but that is another matter. ?
>> (2) I think the semantics of atom are wrong in the draft. It would
>> have helped somewhat to consult on the work of the atom-owl [1]
>> group before putting this spec together.
>>
>> The key problem is the value of the subject:
>>
>> [[
>> <atom:entry>
>> <atom:id>http://example.com/a</atom:id>
>> <atom:title>Test</atom:title>
>> </atom:entry>
>>
>> and the map above as a child of at:entrymap, the following triple
>> would be implied;
>>
>> <http://example.com/a> <http://purl.org/dc/elements/1.1/title>
>> "Test" .
>> ]]
>>
>>
>> By making the value of the subject be the resource identified by
>> the id, you are going against what at least two
>> independently built ontologies for atom consider to be the value of
>> the subject of the entry.
>
> That is only the default; see the subject attribute.
Yes but I believe your default is very misleading. Your default breaks
the best understanding one can put together of what the atom semantics
is.
It also breaks a basic intuition of atom syntax which is given by the
6.4.1 section
[[
The element can be interpreted as a simple property (or name/value
pair) of the parent element that encloses it. The pair consisting of
the namespace-URI of the element and the local name of the element can
be interpreted as the name of the property. The character data content
of the element can be interpreted as the value of the property. If the
element is empty, then the property value can be interpreted as an
empty string.
]]
Ie the subject of the element should be the enclosing element, not
another resource inside the element.
>
>
> [...]
>
>> It cannot be therefore that you have correctly captured the
>> semantics of the atom format.
>
> That is not a goal for this effort.
You are trying to entice people to use rdf in atom. So clearly the
idea here is that atom has a semantics, otherwise the use of rdf in
atom would be useless. The extensions you propose will create a lot of
confusion in the rdf and atom camps as what everyone is speaking
about. But that may well be your main aim.
Henry
>
>
> Cheers,
>
>
> --
> Mark Nottingham http://www.mnot.net/