headers
Bob Ferris | 6 Feb 2011 21:26
Gravatar

Re: [foaf-dev] FOAF-RDF-2-(X)HTML+RDFa-izer?

Am 06.02.2011 21:13, schrieb Bob Ferris:
> Hi Michael,
>
> Your project idea sounds interesting and the example page looks really
> good. However, I can't find any hypertext link that references your FOAF
> profile in a RDF serialization format. With RDFa the content negotiation
> step is primarily not necessary. This enables an easier deployment. No
> configuration for content negotiation is necessary. Although, it could
> still be deployed and the (X)HTML+RDFa lives side by side with the other
> RDF serialization formats. Whereby, the (X)HTML+RDFa document has the
> advantage that it is out-of-the-box usable by human and machine agents.
> It enables me to serve processable views for both humans and machines in
> one document.
> I think (X)HTML+RDFa deployment is crucial for FOAF profiles and for
> Linked Data deployment in general (see for example GoodRelations).

Btw, I saw your weighted skills in the profile. I co-developed recently 
the Cognitive Characteristics Ontology[1] which should exactly address 
such a modelling. That means, it enables you to describe cognitive 
pattern e.g., skills, expertises or interests, simply in form of a 
triple or extended. An extended version can include for instance a 
weighting of the cognitive pattern (by utilizing the Weighting 
Ontology[2]). Please have a look at the examples in the Cognitive 
Characteristics Ontology specification documentation e.g., this one [3].
When applying a weighting it is especially important that the scale 
should be no out-of-band knowledge. It should rather be possible to 
discover that knowledge as needed. Hence, it is useful to also provide 
scales in machine processable form.

Cheers,
(Continue reading)

Permalink | Reply |
headers
Leif Warner | 6 Feb 2011 22:16
Picon
Gravatar

Re: [foaf-dev] FOAF-RDF-2-(X)HTML+RDFa-izer?

If you want manual control over that sort of thing, there's XSPARQL that allows you to write SPARQL queries that return XHTML.


Also, more simply, I've been working on using RDFa as a templating / data-binding language for a sort of web framework. [1]
The basic idea is that this:
<p xmlns:foaf="http://xmlns.com/foaf/0.1/">
   <span property="foaf:name"/>
   <a href="" rel="foaf:knows"><span property="foaf:name"/></a>
</p>

gets filled out to this:

<p xmlns:foaf="http://xmlns.com/foaf/0.1/">
 <span property="foaf:name">Leif</span>
 <a href="http://john.com" rel="foaf:knows"><span property="foaf:name">John</span></a>
 <a href="http://bill.com" rel="foaf:knows"><span property="foaf:name">Bill</span></a>
</p>

The idea is just that "property" lists properties of the current subject, and "rel" traverses to a new subject, filling out an "href" or "resource" or whatever along the way.  Runs against Jena right now, but that could easily be changed.

It's quite similar to what's used in Callimachus, a well-developed project, it turns out. [2]

-Leif Warner


On Sun, Feb 6, 2011 at 12:26 PM, Bob Ferris <zazi-o0kiOr3Ba/rk1uMJSBkQmQ@public.gmane.org> wrote:
Am 06.02.2011 21:13, schrieb Bob Ferris:
> Hi Michael,
>
> Your project idea sounds interesting and the example page looks really
> good. However, I can't find any hypertext link that references your FOAF
> profile in a RDF serialization format. With RDFa the content negotiation
> step is primarily not necessary. This enables an easier deployment. No
> configuration for content negotiation is necessary. Although, it could
> still be deployed and the (X)HTML+RDFa lives side by side with the other
> RDF serialization formats. Whereby, the (X)HTML+RDFa document has the
> advantage that it is out-of-the-box usable by human and machine agents.
> It enables me to serve processable views for both humans and machines in
> one document.
> I think (X)HTML+RDFa deployment is crucial for FOAF profiles and for
> Linked Data deployment in general (see for example GoodRelations).

Btw, I saw your weighted skills in the profile. I co-developed recently
the Cognitive Characteristics Ontology[1] which should exactly address
such a modelling. That means, it enables you to describe cognitive
pattern e.g., skills, expertises or interests, simply in form of a
triple or extended. An extended version can include for instance a
weighting of the cognitive pattern (by utilizing the Weighting
Ontology[2]). Please have a look at the examples in the Cognitive
Characteristics Ontology specification documentation e.g., this one [3].
When applying a weighting it is especially important that the scale
should be no out-of-band knowledge. It should rather be possible to
discover that knowledge as needed. Hence, it is useful to also provide
scales in machine processable form.

Cheers,


Bob


PS: If have further questions re. the Cognitive Characteristics
Ontology, don't hesitate to ask me ;)


[1] http://purl.org/ontology/cco/core#
[2] http://purl.org/ontology/wo/core#
[3]
http://purl.org/ontology/cco/cognitivecharacteristics.html#sec-cognitive-patterns-example

_______________________________________________
foaf-dev mailing list
foaf-dev@...
http://lists.foaf-project.org/mailman/listinfo/foaf-dev
Permalink | Reply |

Gmane