Re: Pragmatic Authoring Tools for RDF Stores

Hi Steven,

On 09/22/2018 03:36 PM, Steven Harms wrote:
> Questions:
> 
> 1. Is authoring like I described above desirable, expected? Is it hoped 
> that individuals would think "Hm, instead of a bulleted list in a Google 
> Doc, I'll add these notes using (sought RDF tool)." Or is the 
> expectation that some other storage / interface / solution that emits 
> RDF will be the primary interface?

I agree with you that it is desirable to have good tools for end-user
RDF authoring. In fact, most existing RDF datasets are programmatically
generated from existing data sources in other formats (e.g. DBpedia is
generated from Wikipedia pages). Whether this is so because of a lack
of tools for manual authoring or the other way around, I don't know.

There is a well-known authoring tool, Protégé, but it requires good
understanding of the underlying logic and it is more for ontology
authoring than for ontology/schema population.

> 2. Assuming individual authorship /is/ a desired thing, is the 
> "Interface" I described above a reasonable sketch of the user experience?
> 
> 3. Assuming individual authorship as desirable and the UX as 
> appropriate, why doesn't that (seem to?) exist?

In my research group (SemLIS, www-semlis.irisa.fr), we find
individual authorship very desirable, and we have done some
work on it. See in particular the following paper, which
provides an illustration on the manual RDF annotation of
comics panels.

Hermann, Alice, Sébastien Ferré, and Mireille Ducassé.
"An interactive guidance process supporting consistent updates of RDFS 
graphs."
International Conference on Knowledge Engineering and Knowledge 
Management. Springer, Berlin, Heidelberg, 2012.
https://hal.inria.fr/hal-00757083/file/EKAW2012_HermannFerrA_DucassA_.pdf

We have a prototype platform SEWELIS at 
http://www.irisa.fr/LIS/ferre/sewelis-servolis/ where you can create a 
RDF store, control who has access to
it, import RDF files, author new RDF descriptions, and query the
data, all without being exposed directly to the formal languages.

> While I doubt that RDF Collections will be used for casual 
> shopping-lists etc., there are a number of places where they ought be 
> useful in both home and business (e.g. `home:photoAlbum2009A 
> schema:about lifeEvents:tripToLondon`) and being able to author them in 
> a pleasant UI that enriches a collection in a virtuous cycle seems to be 
> the future this group is striving toward.

The virtuous circle thing is what we tried to obtain by providing
guidance during the authoring process based on existing data.
If you try SEWELIS, I recommend either to import existing triples as
examples to feed the guidance, or to create the schema on the fly
while creating instances. Feedbacks welcome!


I can also point to "Semantic Forms" at 
http://www.semantic-forms.cc:9111/ developed and maintained by Jean-Marc 
Vanel.

Sébastien

Received on Saturday, 22 September 2018 15:12:19 UTC