R: summary of today's telco

Hi Aldo,

point 4) originates from the intention to make the metadata module able to describe any kind of encoded lexical content (not only expressed through Lemon). To this end, it is possible in Lime to represent also lexicalizations provided by rdfs, skos, skosxl labeling properties, by asserting their respective vocabulary URIs as object of the property: lime:lexicalizationModel.
Furthermore, if an ontology has no lexicalization at all, but its entity URIs are meaningful in some way (e.g. CamelCase  terms) we can state that by putting a special URI (again, as object of the property: lime:lexicalizationModel) to represent this.
The temporary URI was: http://uri4uri.net/vocab (I proposed this initially just as a placeholder) but I guess something more explicit, such as, http://www.w3.org/ns/lemon/meaningfulURIs would be more..meaningful itself ;)

This possibility arose some discontent as it could seem like encouraging developers to use evocative URIs and to forget proper lexicalizations. On the other hand metadata, in doing its job, should really represent reality and not only ideal situations, and we all know how many ontologies are available without any label at all, and with meaningful URIs. Thus Philipp proposed this compromise: keep this URI out from the description of lexicalizationModel, and have a small paragraph in the specification discouraging the use of (only) meaningful URIs but allowing to represent their use as a lexicalization in Lime.

Hope it clarifies and properly reports our discussion on the telco on this point,

Cheers,

Armando


________________________________
Da: Aldo Gangemi<mailto:aldo.gangemi@gmail.com>
Inviato: ‎18/‎07/‎2015 09:29
A: Philipp Cimiano<mailto:cimiano@cit-ec.uni-bielefeld.de>
Cc: public-ontolex@w3.org<mailto:public-ontolex@w3.org>
Oggetto: Re: summary of today's telco

Hi Philipp thanks. I do not understand (4): what is the intended scope of
discouraging from using URIs to encode linguistic knowledge? An example?
Aldo

On Friday, July 17, 2015, Philipp Cimiano <cimiano@cit-ec.uni-bielefeld.de>
wrote:

>  Dear all,
>
>  we took the following decisions today during the final ontolex telco:
>
> 1) The lexicalization model for ontolex should be:
> http://www.w3.org/ns/lemon/all
>
> This concerns especially lime/example2
>
> 2) The property Average Number of Conceptualizations will be renamed into
> avgAmbiguity
>
> 3) The property Average Number of Lexical Entries per Concept will be
> renamed into avgSynonymy
>
> 4) We will add a sentence discouraging people from enconding linguistic
> knowledge in URIs, but will also point to the fact that if they do so, they
> should at least provide information on this as metadata, in particular
> including the language. For this, the property http://uri4uri.net/vocab
> can be used as value of the property lexicalizationModel. In fact, we
> should strongly encourage people to use rdfs:label as lexicalization at the
> very least.
>
> 5) In the properties counting number of lexicalizations per ontology
> entity, etc. we should make clear that the entities of the ontology that
> constitute metadata should not be counted.
>
> 6) The domain of referenceDataset should be extended to include
> LexicalLinkSet
>
> 7) The domain of references should be extended to include LexicalLinkSet
>
> 8) The domain of concepts should be extended to include
> ConceptualizationSet
>
> 9) We decided to keep LexicalSense and OntoMap as a subclass of Lexical
> Sense, while realizing that this is possibly not the best choice. I would
> have favoured having the OntoMap as a separate entity that points to the
> LexicalSense, but so be it...
>
> 10) We discussed what to do with the definition property and decide to
> remove it and recommend in examples to use the skos:definition property.
>
> 11) We discussed the confidence property. There was a majority in favour
> of introducing the confidence property in vartrans. As a compromise, we
> agreed to introduce it into linginfo.
>
> The corresponding examples need to be update.
>
> OK, I think this is all for now...
>
> Armando/Manuel agreed to implement changes 1-9
>
> John agreed to add the property confidence into linginfo
>
> I will rework all the examples as I finally go over the whole spec...
>
> Kind regards,
>
> Philipp.
>
> --
> --
> Prof. Dr. Philipp Cimiano
> AG Semantic Computing
> Exzellenzcluster für Cognitive Interaction Technology (CITEC)
> Universität Bielefeld
>
> Tel: +49 521 106 12249
> Fax: +49 521 106 6560
> Mail: cimiano@cit-ec.uni-bielefeld.de <javascript:_e(%7B%7D,'cvml','cimiano@cit-ec.uni-bielefeld.de');>
>
> Office CITEC-2.307
> Universitätsstr. 21-25
> 33615 Bielefeld, NRW
> Germany
>
>