is ResourceMap needed?

[sheer]

Hi all,

I don not understand why ResourceMap as a concept is needed? is not
the Aggregation resource sufficient?
why do I need to be redirected to another resource of different type
to get the description of an Aggregation? is not it possible to get
such description by dereferencing the URI of the Aggregation itself?

I am not an expert, but in my point of view that is not exactly what
is described in Linked Data Tutorial which is mentioned as a
foundation of the OAI-ORE specification.

Thanks in advance for any help...

Fadi Maali

[Robert Sanderson]

Hi Fadi,

The Aggregation is a non-information resource, a concept. It's the
concept of the collection, rather than an explicit serialization of
the graph that describes that collection.

For an analogue, you have a shopping list in your head that contains 5
items (the aggregation). You can write down that list in different
ways (the different resource maps).
Now think about 'creation' timestamp for each. The Aggregation was
created when you thought of the things you needed to buy. You then
created a serialization when you wrote it down (so a different time to
when you thought of the list), and then maybe you wrote it down again
on a different piece of paper (another different time). In order to
have these timestamps all attached to the right resources, there needs
to be different URIs for them... and hence we need a URI for the
Aggregation, and one for each Resource Map.

As an example of linked data doing this, consider
http://dbpedia.org/resource/Paris. This is a URI for the city of
Paris, France. It then redirects you to
http://dbpedia.org/page/Paris, which is an HTML description of the
resource. Or to http://dbpedia.org/data/Paris.rdf, which is the
rdf/xml description.

Hope that helps.

Rob

> --
> You received this message because you are subscribed to the Google Groups "OAI-ORE" group.
> To post to this group, send email to oai...@googlegroups.com.
> To unsubscribe from this group, send email to oai-ore+u...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/oai-ore?hl=en.
>
>

[Bernhard Haslhofer]

Hi,

as I understand, the main difference between Linked Data and the OAI-ORE approach is that an OAI-ORE model provides explicit links to serializations of data (ResourceMaps). Linked Data sources usually do not explicitly link a resource (http://dbpedia.org/resource/Paris) with its serializations (http://dbpedia.org/data/Paris.rdf, http://dbpedia.org/page/Paris) but leaves the task of delivering the "appropriate" information resource to the server hosting the data - usually this is done by content negotiation.

Best,
Bernhard

On Mar 11, 2010, at 8:06 PM, Robert Sanderson wrote:

> Hi Fadi,
>
> The Aggregation is a non-information resource, a concept. It's the
> concept of the collection, rather than an explicit serialization of
> the graph that describes that collection.
>
> For an analogue, you have a shopping list in your head that contains 5
> items (the aggregation). You can write down that list in different
> ways (the different resource maps).
> Now think about 'creation' timestamp for each. The Aggregation was
> created when you thought of the things you needed to buy. You then
> created a serialization when you wrote it down (so a different time to
> when you thought of the list), and then maybe you wrote it down again
> on a different piece of paper (another different time). In order to
> have these timestamps all attached to the right resources, there needs
> to be different URIs for them... and hence we need a URI for the
> Aggregation, and one for each Resource Map.
>
> As an example of linked data doing this, consider
> http://dbpedia.org/resource/Paris. This is a URI for the city of

> Paris, France. It then redirects you tohttp://dbpedia.org/page/Paris, which is an HTML description of the

______________________________________________________
Research Group Multimedia Information Systems
Department of Distributed and Multimedia Systems
Faculty of Computer Science
University of Vienna

Postal Address: Liebiggasse 4/3-4, 1010 Vienna, Austria
Phone: +43 1 42 77 39635 Fax: +43 1 4277 39649
E-Mail: bernhard....@univie.ac.at
WWW: http://www.cs.univie.ac.at/bernhard.haslhofer

[fadi maali]

Hi,

I do agree with Bernhard that this is a "difference" between  Linked Data and OAI-ORE. I read somewhere that ResourceMap used to be a subclass of NamedGraph which "justifies" its existence although information can still be attached to the retrieved document (<> dct:modified "20-02-2010".) and each document can be treated as named graph if to be aggregated in a triple store...

I still think that having a dedicated resource (a.k.a a ResourceMap) to provided description of "another" resource (a.k.a Aggregation) might not be necessary. As I see it, many constraints in the model can be automatically guaranteed if the description of Aggregation is obtained by dereferencing it and without the use of ResourceMap. As examples:

1. A Resource Map MUST describe a single Aggregation

2. There MUST be at least one Authoritative Resource Map for each Aggregation

That also means minting/maintaining less URIs.

Thanks and regards,

Fadi Maali

[Herbert Van de Sompel]

On Mar 12, 2010, at 3:37, fadi maali <fadi...@gmail.com> wrote:

I still think that having a dedicated resource (a.k.a a ResourceMap) to provided description of "another" resource (a.k.a Aggregation) might not be necessary. 

If an Aggregation would be a document you would be right. But it's not. It is a set, which is a conceptual construct, an abstraction. In that way it is a non-information resource. It is described by an information resource that is the ResourceMap. Two different resources, One a concept the other a document describing the concept. Not something ORE invented, just following guidance from those that led the way re semantic web and linked data. See, for example Cool URIs for the Semantic Web.

Cheers

Herbert 

[Robert Sanderson]

Hi Fadi,

On Fri, Mar 12, 2010 at 10:37 AM, fadi maali <fadi...@gmail.com> wrote:
> I do agree with Bernhard that this is a "difference" between  Linked Data
> and OAI-ORE. I read somewhere that ResourceMap used to be a subclass of
> NamedGraph which "justifies" its existence although information can still be
> attached to the retrieved document (<> dct:modified "20-02-2010".) and each
> document can be treated as named graph if to be aggregated in a triple
> store...

If you think about this a little further, you'll see that in order to
attach a different modified date (or any other metadata) to the
abstract set and to the concrete document, each needs its own URI.

For example, if I create an Aggregation but you create a Resource Map
that describes it, then in order to attach dcterms:creator, the
Aggregation and Resource Map must have different URIs.

To use a dbpedia example, the creator of the city of Paris
(dbpedia.org/resource/Paris) and the creator of the graph that
describes Paris (dbpedia.org/data/Paris.rdf) are very different
people!

Hope that helps,

Rob

[fadi maali]

hi all,

Thanks a lot for your help! completely convinced now :)

Another example of a similar use might be foaf:Person and foaf:PersonalProfileDocument

Regards,

Fadi Maali

--