- From: Danny Ayers <danny.ayers@gmail.com>
- Date: Wed, 27 Jun 2012 22:27:11 +0200
- To: Dan Brickley <danbri@danbri.org>
- Cc: Egor Antonov <elderos@yandex-team.ru>, Greg Grossmeier <greg@creativecommons.org>, Phil Barker <phil@pjjk.net>, Stuart Sutton <sasutton@dublincore.net>, zoe.f.rose@gmail.com, "public-vocabs@w3.org Vocabs" <public-vocabs@w3.org>, Peter Markov <peter-markov@yandex-team.ru>
I do like the style of the-specification and also think it would be great to get this under the schema.org umbrella. But...a couple of actually rather general points - For one, what do these things /mean/? I hate to be a logic Nazi, but is "Expected Type" the same as rdfs:range? I'm not necessarily suggesting using RDFS terminology, but "Expected Type" is rather vague - what if you don't get the expected type, what interpretation should you take? The other point I take a little issue with is the whole pattern around schema.org/URL - In this context: useRightsUrl The URL where the owner specifies permissions for using the resource. That seems contrary to the spirit, and very possibly the practicality of WebArch. Why not just useRights - with the object being the name (URL) of the resource that fulfils..? Maybe nitpicking, but castles built of sand - last for millenia with a bit of cement. Cheers, Danny. On 27 June 2012 21:03, Dan Brickley <danbri@danbri.org> wrote: > +cc Greg, Zoe, Phil, Stuart ... with whom I've variously been > discussing LRMI integration > > On 26 June 2012 17:35, Egor Antonov <elderos@yandex-team.ru> wrote: >> Hi folks! >> Some of our proposals (Medical/Health, TechArticle, LRMI) and also some of >> our Yandex people have common ideas floating around. >> I mean that currently a publisher cannot explicitly specify target audience >> of his site/page/CreativeWork. >> >> Medical vocabulary proposal contains a type MedicalAudience, which splits >> people of medicine into categories expressed as an enumeration. >> >> LRMI proposal contains intendedEndUserRole property with the same purpose, >> contaning some text (which is also a soft enum). > > Looking at this again ... in terms of "what can we do now for > integration, while still moving quickly towards including LRMI 1.0"? > > >From LRMI, as you say, http://www.lrmi.net/the-specification > > * intendedEndUserRole schema.org/Text The individual or group for > which the work in question was produced. Ex: “student” Ex: “teacher” > > Would a minimalistic improvement here be, to allow intendedEndUserRole > to also (as an alternative) point to a thing of type > http://schema.org/Audience ? > > This would immediately -for example- permit use of the specific > defined instances of http://schema.org/MedicalAudience we added > earlier this week such as http://schema.org/Clinician ... as well as > other approaches as explored in > http://www.w3.org/wiki/WebSchemas/Audience ... While schema.org might > define a few useful kinds of audience, we won't be able to anticipate > every such type - hopefully various educational standards and others > will get expressed this way and included via URL references. > > Would an 'intended end user role' of http://schema.org/Clinician even > make sense in an LRMI context? Should we define similar builtins for > the given examples, i.e. Student and Teacher ? Are there controlled > lists out there somewhere? > > How does the general approach sound here, i.e. allowing > intendedEndUserRole to point to an instance of > http://schema.org/Audience? I'm concerned that we find a way to have a > general approach to characterising audiences, but also to move ahead > quickly with LRMI... > > cheers, > > Dan > > ps. for in-depth LRMI background, see videos in > http://www.lrmi.net/metadata-lab-video > -- http://dannyayers.com http://webbeep.it - text to tones and back again
Received on Wednesday, 27 June 2012 20:27:40 UTC