[NLP2RDF] NIF ontology: broken links

[Antonin Delpeuch (lists)]

Hi Sebastian!

Great, that tooling looks very useful indeed! I am neither familiar with
SHACL nor confident with NIF so far so I am probably not best placed to
rewrite NIF in SHACL, but that does seem like a worthwhile project.

Concerning human-readable documentation, I have found this:
https://github.com/dice-group/gerbil/wiki/How-to-generate-a-NIF-dataset
http://brown.nlp2rdf.org/
That's what is closest to what I need, so I will try to improve this as
I find answers to my questions.

Best,
Antonin

On 2/5/19 9:14 PM, Sebastian Hellmann wrote:
> Hi Antonin,
> 
> On 05.02.19 16:02, Antonin Delpeuch (lists) wrote:
>> Hi Sebastian,
>>
>> Great, thanks! Beyond the description of the ontology, it would be quite
>> useful to have a more explanatory description of NIF as a file format
>> for NERD datasets. The ontology only loosely describes the format: for
>> instance, it is not clear to me which statements are required, which
>> ones are optional, which ones can have multiple values, and so on.
> 
> you need to cut me some slack here. We really tried to do that back then
> with the validation model in OWL. However, OWL was not well-suited, so
> my PhD student Dimitris Kontokostas made the W3C standard SHACL to
> directly address this issue:
> 
> * *NLP data cleansing based on Linguistic Ontology constraints
> <http://jens-lehmann.org/files/2014/eswc_rdfunit_nlp.pdf> /by/ Dimitris
> Kontokostas, Martin Brümmer, Sebastian Hellmann, Jens Lehmann, and
> Lazaros Ioannidis * /in/ Proc. of the Extended Semantic Web Conference 2014
> 
> * https://www.w3.org/TR/shacl/
> 
>> Moreover, the NIF format also uses predicates from other namespaces
>> which are not described there, so that makes it even harder to
>> understand, I think.
> 
> This is also a common problem, which can be addressed by SHACL as well.
> 
> If I were to do NIF again, I would only use SHACL and RDFS and not OWL
> at all. SHACL is a much better modelling language and you can scope over
> external classes.
> 
>> Having such a document (with examples) would really encourage people
>> from outside the semantic web community (such as NLP researchers not
>> familiar with these terminologies) to embrace the format, I think.
> 
> RDFUnit (https://github.com/AKSW/RDFUnit) can be hosted as a webservices
> with SHACL, so you will get a free validator, see e.g. here:
> 
> http://databus.dbpedia.org:8080/shacl/validate?d=http://kurzum.net/webid.ttl&A&s=https://raw.githubusercontent.com/dbpedia/webid/master/voc/webid-shacl.ttl
> 
> If you have the time, I think it would be good to rewrite NIF in SHACL
> and then publish all versions on the Databus.  Things will get much
> easier for everybody.
> 
> 
> All the best,
> 
> Sebastian
> 
> 
> 
> 
>> Best,
>> Antonin
>>
>> On 2/5/19 2:44 PM, hellmann at informatik.uni-leipzig.de wrote:
>>> Hi Antonin,
>>>
>>> The turtle file is here
>>> http://persistence.uni-leipzig.org/nlp2rdf/ontologies/nif-core/nif-core.ttl
>>>
>>> There is a broken internal redirect pending to be fixed.
>>>
>>> 2.1 is here https://github.com/NLP2RDF/ontologies
>>>
>>>
>>> Over the years it became clear that publishing ontologies like this is
>>> very unpractical. We are working on https://databus.dbpedia.org .
>>>
>>> When it is ready, I will migrate the ontology there. It like maven
>>> central for data, so you can build your app on the right version...
>>>
>>> Sebastian
>>>
>>>
>>>
>>> On February 5, 2019 2:32:29 PM GMT+01:00, "Antonin Delpeuch (lists)"
>>> <lists at antonin.delpeuch.eu> wrote:
>>>
>>>     Hi,
>>>
>>>     I was looking for the definition of the NIF format. It seems that it
>>>     used to be available at:
>>>     http://persistence.uni-leipzig.org/nlp2rdf/ontologies/nif-core#
>>>
>>>     This link currently gives an HTTP 403 error for me. I found it in the
>>>     "Integrating NLP using Linked Data paper":
>>>     http://svn.aksw.org/papers/2013/ISWC_NIF/public.pdf
>>>     Given that this link appears in a published paper, I think it is very
>>>     important that there is an effort to maintain it (otherwise the
>>>     "persistence" in the domain name is not really deserved).
>>>
>>>     The same link is also used at:
>>>     http://persistence.uni-leipzig.org/nlp2rdf/
>>>
>>>     Thanks to Wikidata (https://www.wikidata.org/wiki/Q31794099) I was able
>>>     to find
>>>     http://persistence.uni-leipzig.org/nlp2rdf/ontologies/nif-core/nif-core.html,
>>>     which seems to be about NIF 2.0, what about NIF 2.1?
>>>
>>>     Best regards,
>>>     Antonin Delpeuch
>>>     ------------------------------------------------------------------------
>>>     NLP2RDF mailing list
>>>     NLP2RDF at lists.informatik.uni-leipzig.de
>>>     http://lists.informatik.uni-leipzig.de/mailman/listinfo/nlp2rdf
>>>
>>>
>>> -- 
>>> Sent from my Android device with K-9 Mail. Please excuse my brevity.
> -- 
> All the best,
> Sebastian Hellmann
> 
> Director of Knowledge Integration and Linked Data Technologies (KILT)
> Competence Center
> at the Institute for Applied Informatics (InfAI) at Leipzig University
> Executive Director of the DBpedia Association
> Projects: http://dbpedia.org, http://nlp2rdf.org,
> http://linguistics.okfn.org, https://www.w3.org/community/ld4lt
> <http://www.w3.org/community/ld4lt>
> Homepage: http://aksw.org/SebastianHellmann
> Research Group: http://aksw.org