Named graphs?

[Erik Hetzner]

Hi -

I am reading through the 1.0 OAI-ORE specs. Could somebody explain to
me what happened to named graphs? Thanks.

best,
Erik Hetzner

[Robert Sanderson]

In my opinion, Named Graphs went away because of lack of mainstream
support. GIven the success of the Linked Data recommendations, being
full Named Graphs would be a significant barrier to adoption.

The advantages of Named Graphs really didn't amount to much, compared
to the advantages of the Linked Data style.

Rob

> ;; Erik Hetzner, California Digital Library
> ;; gnupg key id: 1024D/01DB07E3
>
>

[Erik Hetzner]

At Fri, 18 Dec 2009 20:34:37 +0000,

Robert Sanderson wrote:
>
> In my opinion, Named Graphs went away because of lack of mainstream
> support. GIven the success of the Linked Data recommendations, being
> full Named Graphs would be a significant barrier to adoption.
>
> The advantages of Named Graphs really didn't amount to much, compared
> to the advantages of the Linked Data style.

Hi Rob -

I’m a little confused about the relationship of named graphs as
compared with linked data. I will have to look into this more.

According to [1], ore:ResourceMap is a subclass of rdfg:Graph. Is this
a holdover from the named graph stuff?

best,
Erik Hetzner

1. http://www.openarchives.org/ore/1.0/vocabulary

[Robert Sanderson]

We wanted to say that the Resource Map is a graph, and Named Graphs
were the only way to do it that we found, as RDF has only one, global
graph by default.

I wouldn't read too much in to the typing there!

Rob

[Erik Hetzner]

At Fri, 18 Dec 2009 22:20:04 +0000,

Robert Sanderson wrote:
>
> We wanted to say that the Resource Map is a graph, and Named Graphs
> were the only way to do it that we found, as RDF has only one, global
> graph by default.
>
> I wouldn't read too much in to the typing there!

OK, thanks, I won’t.

I presume that the URI of the Resource Map would be the appropriate
identifier to use for a named graph if we were using OAI-ORE in a
named graph context. Is that correct?

Also, I am looking at using OAI-ORE with versioned resources. Does
anybody have any references on the use of OAI-ORE with versioned
resources? I notice that the RDF/XML serialization includes some
examples using dcterms:hasVersion, so I will probably be starting from
there.

best,
Erik Hetzner

[Herbert Van de Sompel]

On Dec 18, 2009, at 4:25 PM, Erik Hetzner wrote:

At Fri, 18 Dec 2009 22:20:04 +0000,
Robert Sanderson wrote:

We wanted to say that the Resource Map is a graph, and Named Graphs

were the only way to do it that we found, as RDF has only one, global

graph by default.

I wouldn't read too much in to the typing there!

OK, thanks, I won’t.

I presume that the URI of the Resource Map would be the appropriate
identifier to use for a named graph if we were using OAI-ORE in a
named graph context. Is that correct?

Actually, that is totally correct. If you are looking for a Named Graph in ORE, then it is the Resource Map.  To add to Rob's feedback, one of the reasons we removed the references to Named Graphs in the ORE specs is that we received feedback that the specs already contained a lot of new concepts for many people: information resources, non-information resources, HTTP 303s, Linked Data, etc. In addition to that, we felt that the explicit reference to the Named Graph work didn't really add anything to the spec or to its understanding. See, the Linked Data Tutorial document also has a clear lineage to the Named Graph work (Chris Bizer was closely involved in both), and that document doesn't reference Named Graphs either. 

Also, I am looking at using OAI-ORE with versioned resources. Does
anybody have any references on the use of OAI-ORE with versioned
resources? I notice that the RDF/XML serialization includes some
examples using dcterms:hasVersion, so I will probably be starting from
there.

I guess you mean versioned Aggregated Resources, not versioned Aggregation?This paper has an example of Aggregations that tie versions together:

http://arxiv.org/abs/0906.2549

Cheers

Herbert

best,
Erik Hetzner

;; Erik Hetzner, California Digital Library
;; gnupg key id: 1024D/01DB07E3

==

Herbert Van de Sompel

Digital Library Research & Prototyping

Los Alamos National Laboratory, Research Library

http://public.lanl.gov/herbertv/

tel. +1 505 667 1267 

[Erik Hetzner]

At Fri, 18 Dec 2009 17:33:09 -0700,

Herbert Van de Sompel wrote:
> Actually, that is totally correct. If you are looking for a Named
> Graph in ORE, then it is the Resource Map. To add to Rob's feedback,
> one of the reasons we removed the references to Named Graphs in the
> ORE specs is that we received feedback that the specs already
> contained a lot of new concepts for many people: information
> resources, non-information resources, HTTP 303s, Linked Data, etc. In
> addition to that, we felt that the explicit reference to the Named
> Graph work didn't really add anything to the spec or to its
> understanding. See, the Linked Data Tutorial document also has a clear
> lineage to the Named Graph work (Chris Bizer was closely involved in
> both), and that document doesn't reference Named Graphs either.

Thanks for the extended explanation. Thinking about it, it makes sense
(if I am not mistaken): you don’t need named graphs until you start
bringing together different resource maps for the same aggregation,
which is an issue not covered by OAI-ORE explicitly.

> I guess you mean versioned Aggregated Resources, not versioned
> Aggregation?This paper has an example of Aggregations that tie
> versions together:
>
> http://arxiv.org/abs/0906.2549

Yes, that is exactly what I was thinking. Thank you for the reference.

best,
Erik Hetzner

[MichaelNelson]

On Dec 18, 8:13 pm, Erik Hetzner <erik.hetz...@ucop.edu> wrote:

> Thanks for the extended explanation. Thinking about it, it makes sense
> (if I am not mistaken): you don’t need named graphs until you start
> bringing together different resource maps for the same aggregation,
> which is an issue not covered by OAI-ORE explicitly.
>

We don't have examples of it in the current docs (again, it seemed
like a lot already),
but the hooks are there with the authoritative vs. non-authoritative
ReMs concept:

http://www.openarchives.org/ore/1.0/datamodel.html#Resource_Map

So I could create a different ReM (URI-R-MLN) that ore:describes your
Aggregation
(URI-A). But dereferencing URI-A would lead to URI-R-EH, not URI-R-
MLN. So anyone
would encountering URI-R-MLN would have to decide which ReM they like/
trust
better. My ReM might actually be "better" by some metric (a fan boy
ReM?),
but it would still be non-authoritative.

regards,

Michael

> best,
> Erik Hetzner

[Erik Hetzner]

At Fri, 18 Dec 2009 21:10:59 -0800 (PST),

Thanks for the pointer, Michael. This makes sense.

best,
Erik