- From: Bernard Vatant <bernard.vatant@mondeca.com>
- Date: Fri, 21 Feb 2014 11:22:38 +0100
- To: jean delahousse <delahousse.jean@gmail.com>
- Cc: Kevin Ford <kefo@3windmills.com>, SKOS <public-esw-thes@w3.org>
- Message-ID: <CAK4ZFVEKGs6hdkMG3uKDVYObJ-JgKkxOiCWn7Rk-_k9_zNM4sw@mail.gmail.com>
Jean One answer is called ... named graphs. Create a named graph for each Concept Scheme, with a different notation in each one. Therefore the unicity of notation is valid inside each named graph. This way you can also add comments, definitions specific to each context, or even modify the hierarchy of concepts (like not taking some branch, or extending the hierarchy to other concepts). The broader-narrower relationships will also be specific to a named graph etc. It does not prevent to specify the datatype for notation, but this is only a syntactic constraint. Hope that helps Bernard 2014-02-21 10:52 GMT+01:00 jean delahousse <delahousse.jean@gmail.com>: > Dear Kerin, > > Thanks for your answer about skos:notation, I crossed the same need at the > Publications Office of the European Union and I prefer your solution as the > one I was thinking of (create sub-properties of skos:notation). > > You might have an answer to this related question: > > The same concept (same URI) is used in a thesaurus and then reused in an > other conceptScheme representing a taxonomy. In each conceptScheme we want > to give an id to the concept related to its location in the conceptScheme > hierarchy, something like 4.53 in the first conceptScheme and 2.4.55 in the > second. We use skos:notation for this property. > > We need to say that one skos:notation is valid for the first conceptScheme > and the second one for the second conceptScheme. > > Would you use the "rdf:datatype" solution to differentiate the 2 > skos:notation and relate them to the proper conceptScheme ? Would you use > an other solution ? > > The reuse of concept (using a single URI, not skos:match) in several > conceptScheme brings a lot of questions, but this is one for which I don't > have any good answer today. > > Yours > Jean > > > > > > 2014-02-17 1:02 GMT+01:00 Kevin Ford <kefo@3windmills.com>: > >> Dear Jacob, >> >> When it came time to convert our MARC Organizations dataset into >> RDF/SKOS, we did not want to lose information and granularity we had >> collected and maintained for years. We've formally maintained and >> published multiple codes for each Organization (one code is the "official" >> one; one code represents a normalized form; another possible code - though >> not presently included - is an ISIL [1]. Regardless, the ISIL, if we >> choose to include it, represents yet another perfectly valid code for the >> same Concept). >> >> In any event, we established two different datatypes, one for each code >> (the "official" one and the normalized one). See, for example: >> >> http://id.loc.gov/vocabulary/organizations/dlcmrc.skos.rdf >> >> Yours, >> Kevin >> >> [1] http://biblstandard.dk/isil/ >> >> -- >> Kevin Ford >> Library of Congress >> Washington, DC >> >> >> >> >> >> >> On 02/14/2014 01:53 PM, Voß, Jakob wrote: >> >>> Hi, >>> >>> I am designing a software that uses SKOS data and unsure about how to >>> limit notations. Are multiple notations per concepts actually used? The >>> advise to use custom datatypes is rather unhelpful because RDF has no >>> default mechanism to express information about datatypes. If most concept >>> schemes have only one notation per Concept, I'd rather make this a >>> constraint instead of implementing edge cases that nobody makes use of >>> anyway. See also my question at http://answers.semanticweb. >>> com/questions/26492/are-skos-concepts-with-multiple- >>> notations-actually-used >>> >>> Jakob >>> >>> -- >>> Jakob Voß >>> Verbundzentrale des GBV (VZG) >>> Abteilung Digitale Bibliothek >>> Platz der Göttinger Sieben 1 >>> 37073 Göttingen >>> Telefon: (49)551 39-10242 >>> Internet: www.gbv.de >>> >>> >> > > > -- > Jean Delahousse > > > ---------------------------------------------------------------------------------------------------- > delahousse.jean@gmail.com - +33 6 01 22 48 55 > http://fr.linkedin.com/in/jeandelahousse > > > -- *Bernard Vatant* Vocabularies & Data Engineering Tel : + 33 (0)9 71 48 84 59 Skype : bernard.vatant http://google.com/+BernardVatant -------------------------------------------------------- *Mondeca* 3 cité Nollez 75018 Paris, France www.mondeca.com Follow us on Twitter : @mondecanews <http://twitter.com/#%21/mondecanews> ----------------------------------------------------------
Received on Friday, 21 February 2014 10:23:27 UTC