- From: Nathan <nathan@webr3.org>
- Date: Wed, 18 Nov 2009 18:00:32 +0000
- To: Aaron Rubinstein <arubinst@library.umass.edu>
- CC: Niklas Lindström <lindstream@gmail.com>, pedantic-web@googlegroups.com, Kingsley Idehen <kidehen@openlinksw.com>, Linked Data community <public-lod@w3.org>
Aaron Rubinstein wrote: > > Niklas Lindström wrote: >> On Wed, Nov 18, 2009 at 5:10 PM, Nathan <nathan@webr3.org> wrote: >>> Kingsley Idehen wrote: >>>> Niklas Lindström wrote: >>>>> Hi Nathan! >>>>> >>>>> >>>>>> So let's say I run an article of content through called >>>>>> "Deforestation >>>>>> and Competing Water Uses" >>>>>> >>>>>> the main subjects of the article are: >>>>>> >>>>>> + http://dbpedia.org/resource/Deforestation >>>>>> + http://dbpedia.org/resource/Reforestation >>>>>> + http://dbpedia.org/resource/Soil_conservation >>>>>> + http://dbpedia.org/resource/Silt (siltation) >>>>>> + http://dbpedia.org/resource/Aberdare_Range >>>>>> + http://dbpedia.org/resource/Tana_River_%28Kenya%29 >>>>>> >>>>>> which is fine, they are dc:subject / foaf:topic etc >>>>>> >>>>> Sounds good. >>>>> >>>>> >>>>>> but then the article is under the general topics of: >>>>>> + http://dbpedia.org/resource/Adaptation_to_global_warming >>>>>> + http://dbpedia.org/resource/Kenya >>>>>> >>>>>> and it mentions: >>>>>> + http://dbpedia.org/resource/Sustainable_forest_management >>>>>> + http://dbpedia.org/resource/Afforestation >>>>>> + http://dbpedia.org/resource/Hydropower >>>>>> + http://dbpedia.org/resource/Municipal_water_supply >>>>>> + http://dbpedia.org/resource/Life_span >>>>>> + http://dbpedia.org/resource/Soil >>>>>> + http://dbpedia.org/resource/Forestry >>>>>> + http://dbpedia.org/resource/Plant >>>>>> + http://dbpedia.org/resource/Reservoir >>>>>> >>>>>> From the aspect of the seeker, these "mentions" are invaluable - if I >>>>>> was doing a report on issues affection reservoirs in kenya, then this >>>>>> data is most valuable and thus related. >>>>>> >>>>>> so, which ontology is most suited for this case of "mentions"? >>>>>> it's not >>>>>> a subject or a tag, and don't want to identify the data as such as >>>>>> that >>>>>> is misleading and could be easily misrepresented within UIs if it >>>>>> were. >>>>>> >>>>> How about dct:references <http://purl.org/dc/terms/references>? >>>>> Defined as "A related resource that is referenced, cited, or otherwise >>>>> pointed to by the described resource.", I figure it is about as >>>>> generic as any "unlabelled" hypertext link. >>>>> >>>>> Best regards, >>>>> Niklas >>>>> >>>>> >>>> Ah! yes, that's even better that my sioc:links_to and foaf:topic >>>> suggestions :-) >>>> >>> perhaps dct:relation is more suited? >>> >>> dct:Relation >>> A related resource. >>> >>> dct:References >>> A related resource that is referenced, cited, or otherwise pointed to by >>> the described resource. >>> >>> thoughts? >> >> Ah, yes, better still! The dct:relation is a subPropertyOf >> dct:references, so that seems to be the "most" generic one. And Dublin >> Core is nicely generic, so I doubt you'd imply anything unintended by >> using dct:relation. >> >> Best regards, >> Niklas >> > > I believe it's the other way around, which would make dct:references > ideal in every way ;) but if you need to distinguish between relations with a citation, and those without then using each in its appropriate place may make sense.. ?
Received on Wednesday, 18 November 2009 18:01:36 UTC