- From: Bernard Vatant <bernard.vatant@mondeca.com>
- Date: Tue, 21 Oct 2008 18:28:40 +0200
- To: Richard Cyganiak <richard@cyganiak.de>
- CC: dbpedia-discussion@lists.sourceforge.net, SKOS <public-esw-thes@w3.org>
Richard Cyganiak a écrit : > Bernard, > > I don't know about plans regarding the namespace. Just one thing: > Wikipedia categories are not transitive, and applying transitivity > inference would produce undesirable results. Suppose so. Maybe I was naively dreaming of some "local transitivity" (say, 2 or 3 steps ...) > This is because the category system is simply not a well-designed > hierarchy, but rather a tagging system where categories can be tagged > as well. Yep. With neither loop nor meta-level consistency control ... > For example, if we applied transitive inference, then category:Berlin > would become a subcategory of category:Mexico, category:Asia, and > category:Jesus. Don't ask why... ;-) Well, I did not figure it was *that* bad ... Makes me wonder if using skos:broader to express dbpedia categories "hierachy" is a good idea at all ... Bernard > > Best, > Richard > > > On 21 Oct 2008, at 12:27, Bernard Vatant wrote: > >> Hi all >> >> I wanted to get at the following data set through >> http://dbpedia.org/sparql >> >> "Churches in Paris, along with French name and description, and >> geo-tagging elements" >> >> The category "Churches in Paris" has typically two levels of >> subcategories, by "Religion", then by "Arrondissement" >> >> So the following query >> >> PREFIX p: <http://dbpedia.org/property/> >> PREFIX dbpedia: <http://dbpedia.org/resource/> >> PREFIX category: <http://dbpedia.org/resource/Category:> >> PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#> >> PREFIX skos: <http://www.w3.org/2004/02/skos/core#> >> PREFIX geo: <http://www.georss.org/georss/> >> >> SELECT DISTINCT ?m ?n ?p ?d >> WHERE { >> ?m rdfs:label ?n. >> ?m skos:subject ?c. >> ?c skos:broader category:Churches_in_Paris. >> ?m p:abstract ?d. >> ?m geo:point ?p >> FILTER ( lang(?n) = "fr" ) >> FILTER ( lang(?d) = "fr" )} >> >> ... yields only the two results which are in a direct subcategory. Which >> means skos:broader transitivity is not supported. OK. >> This leads to several remarks and questions. >> SKOS version declared in the namespace is the 2004 version, which is now >> superseded by the new 2008 SKOS specification and namespace. >> For those not aware of the long debate history in SKOS about broader >> transitivity, the sum up is as following >> >> skos(2004):broader is transitive >> skos(2008):broader is no more transitive. It links only to the direct >> parent. >> skos(2008):broaderTransitive is a superproperty of the above, and it is >> transitive - like skos(2004):broader >> >> To sum it up, the current dbpedia interface declares skos:broader in the >> 2004 namespace, but de facto applies the 2008 semantics (no >> transitivity). >> This is indeed confusing, and given there has been already a lot of >> confusion in people minds about that affair, it's too bad. >> Has dbpedia in project to switch from SKOS 2004 to SKOS 2008 namespace >> and semantics, and if yes when? And in such a case case, will it support >> skos2008:broaderTransitive queries? As a side note, skos:subject is >> deprecated in SKOS 2008. >> >> Back to my churches in Paris, is there currently a workaround to get the >> results for all subcategories in a single query? >> -- *Bernard Vatant *Senior Consultant Vocabulary & Data Engineering Tel: +33 (0) 971 488 459 Mail: bernard.vatant@mondeca.com <mailto:bernard.vatant@mondeca.com> ---------------------------------------------------- *Mondeca** *3, cité Nollez 75018 Paris France Web: www.mondeca.com <http://www.mondeca.com> Blog: Leçons de Choses <http://mondeca.wordpress.com/> ----------------------------------------------------**
Received on Tuesday, 21 October 2008 16:29:43 UTC