|
Having metadata about books, journals etc. we want to foaf:isPrimaryTopicOf # useful, but not specific enough ( look up the namespace prefix ) Also interesting is Karen Coyles "Cheat Sheet", listing link relations in FRBR . Do you know of any more ? |
|
I now added a very general property for linking a resource to its online fulltext to lobid vocab: http://lobid.org/vocab/lobid#fulltextOnline It may be used to directly link a resource to an online fulltext or to link it to a landing page which provides a link to the fulltext. (The problem is that we can't distinguish both cases in the library data we get...) Here's the description in turtle:
|
|
You'd think it by how hard people go looking for existing properties, but it's not a crime to coin your own properties. If you're not 100% satisfied with existing properties or you're having intermitable meetings about which one to use, that's a good reason to stop having meetings, make your own up, and move forward with your project. Note if you or somebody else decides that another property is good for this you can always put this in your T-Box
|
|
We had a discussion about this over at #swig Explanation for the down vote? Rurik, I am in need of properties. Content negotiation would not help because my URIs are about things (the book made out of paper). I want to link the description about that book with i.e. an ePub Version, Scan, OCR whatever. (...I have not downvoted) Sure, the answer is owl:sameAs, perhaps that got lost in the discussion. Content negotiation was a side issue, unrelated to the specific question asked here. However, I suspect that providing the content negotiation as described adds value to this discussion as well as a uri per thing :) owl:sameAs is NOT ok (if it were so simple ...) because it is NOT the same resource (i.e. a "book of paper" is NOT the same as a "PDF"). I am working for a library organisation, and they are really nitpickicking: if two books are principally the same but are different editions they would not treat them as owl:sameAs ... they take a bearing of frbr). Nitpicking about AACR2/RDA & FRBR doesn't equate to nitpicking about data. In any case, since cataloguing is still based on book-in-hand practice, you can safely that each record actually represents a real thing. Represent each subsequent "version" (pdf, epub) as a new URI, link these to a standard title object (not a derived Work-Expression-Manifestation chain) and then use owl:sameAs to link between the "identical" representations (pdf, epub, paper, rdf) of the URIs for each item (which magically becomes a "manifestation"). If you get different editions, link to the standard title. |
|
General enrichment is too fuzzy to find a property but Let I'd model
The question whether hm, I did not grasp that entirely I think, because in my understanding your example gives no answer to what kind of exemplar it is. I wouldn't know if it is a online resource or just another URI for the Real World Object (RWO). Your consideration about OAI-ORE I share, though. |


A little bit confused. You have a URI to identify a book, and you want a property to link it to a digital copy? Correct?
exactly, that is one scenario.
So you want various specific versions of
rdfs:seeAlso/foaf:isPrimaryTopicOf? I think you might need to define them from scratch.Another vocabulary you might want to check out:
which has
dblp:crossref.