Tom, I totally agree that the relationships must be much simpler than 
what is in FRBR -- a small selection from FRBR would suffice. Note that 
the OL has Works and Editions, and that Editions are kind of a 
combination of FRBR Expression and Manifestation. This means that there 
may be some Expression->Expression relationships that could be applied 
to OL Editions.

I have no problem with recording that one work cites another. That to me 
is not terribly different to "adapts" -- although the citation itself is 
a piece of the content, it establishes an intellectual relationship 
between the works that perhaps could be better expressed as "uses ideas 
from" or "builds on."

kc

On 11/18/12 3:05 PM, Tom Morris wrote:
> I like the idea of recording relationships, but I think of citations as
> being content as opposed to metadata which seems to be OpenLibrary's
> main focus.
>
> The FRBR relationships seem like they could be simplified.  Freebase
> uses a single adapted_work/adapted_from property pair without having
> separate properties for each source/target type pairing like FRBR seems
> to have.  That still allows you to link the film adapted from the
> musical adapted from the play adapted from the novel.  Similar
> simplifications could probably be applied to some of the other FRBR
> relationships.
>
> Tom
>
>
> On Sun, Nov 18, 2012 at 5:45 PM, Karen Coyle <[email protected]
> <mailto:[email protected]>> wrote:
>
>     Ben, I don't know of any that have implemented the FRBR relationships
>     *yet*. That leaves the arena wide open, and being the first to do this
>     would be very interesting!
>
>     kc
>
>     On 11/18/12 2:23 PM, Ben Companjen wrote:
>      > Of course, FRBR! I could have thought of that (and CiTO) :)
>      > So the thinking of relationships we might need has already been done
>      > and "only" the thinking of how to record these relationships and
>      > designing and implementing changes to the interface are left.
>      >
>      > Are there other catalogues that show any of the FRBR work
>     relationships?
>      >
>      > On 18 November 2012 17:21, Karen Coyle <[email protected]
>     <mailto:[email protected]>> wrote:
>      >> FRBR provides a set of relationships between works and it would
>     be good
>      >> to have links that are meaningful, not just "link." "Cites"
>     isn't one of
>      >> them, but there are citation vocabularies. If there were more
>     lists in
>      >> OL those could be mined for statistical relationships.
>     Librarything is
>      >> able to do this so they can give a small set of the most related
>     books.
>      >>
>      >> - my "cheat sheet" of FRBR relationships:
>      >> http://kcoyle.net/rda/group1relsby.html
>      >>     it shouldn't be too hard to make a selection of a small
>     number of
>      >> these for OL use
>      >> - the CITO vocabulary:
>      >>
>      >>
>     
> http://speroni.web.cs.unibo.it/cgi-bin/lode/req.py?req=http:/purl.org/spar/cito
>      >>     again, a small selection might be useful
>      >>
>      >> kc
>      >>
>      >>
>      >> On 11/18/12 5:22 AM, Ben Companjen wrote:
>      >>> Hi all,
>      >>>
>      >>> Although I know the developers are mostly working on non-OL
>     projects,
>      >>> I would like to share a (possibly very old) idea for enhancement of
>      >>> Open Library: direct links from one work to another.
>      >>>
>      >>> It is not an original idea, as most digital libraries for
>     scientific
>      >>> literature have links from articles to cited/referenced articles or
>      >>> book chapters. For me as a student it has become a natural way of
>      >>> browsing the "library".
>      >>> Today I wanted to look up the Thesaurus of English words and
>     phrases
>      >>> by Peter Mark Roget and found not only the multiple works (which
>      >>> should most likely be one work) by Roget, but also "Roget's
>     thesaurus
>      >>> ..." by other authors - edited versions, perhaps reviews or
>     manuals,
>      >>> etc.
>      >>>
>      >>> I think being able to say "works A and B are related, because A
>     is (a
>      >>> review of | a manual to | partly based on | ...) B" would be a
>     great
>      >>> addition to Open Library as a catalog. (And I didn't see this
>     idea in
>      >>> the list of issues on GitHub.)
>      >>>
>      >>> Ben
>      >>>
>      >>> P.S. Yes, you could add a hyperlink to the work, but if you
>     used the
>      >>> label to express the relationship between works, it would only make
>      >>> sense to human users.
>      >>> _______________________________________________
>      >>> Ol-tech mailing list
>      >>> [email protected] <mailto:[email protected]>
>      >>> http://mail.archive.org/cgi-bin/mailman/listinfo/ol-tech
>      >>> To unsubscribe from this mailing list, send email to
>     [email protected] <mailto:[email protected]>
>      >>>
>      >>
>      >> --
>      >> Karen Coyle
>      >> [email protected] <mailto:[email protected]> http://kcoyle.net
>      >> ph: 1-510-540-7596 <tel:1-510-540-7596>
>      >> m: 1-510-435-8234 <tel:1-510-435-8234>
>      >> skype: kcoylenet
>      >> _______________________________________________
>      >> Ol-tech mailing list
>      >> [email protected] <mailto:[email protected]>
>      >> http://mail.archive.org/cgi-bin/mailman/listinfo/ol-tech
>      >> To unsubscribe from this mailing list, send email to
>     [email protected] <mailto:[email protected]>
>      > _______________________________________________
>      > Ol-tech mailing list
>      > [email protected] <mailto:[email protected]>
>      > http://mail.archive.org/cgi-bin/mailman/listinfo/ol-tech
>      > To unsubscribe from this mailing list, send email to
>     [email protected] <mailto:[email protected]>
>      >
>
>     --
>     Karen Coyle
>     [email protected] <mailto:[email protected]> http://kcoyle.net
>     ph: 1-510-540-7596 <tel:1-510-540-7596>
>     m: 1-510-435-8234 <tel:1-510-435-8234>
>     skype: kcoylenet
>     _______________________________________________
>     Ol-tech mailing list
>     [email protected] <mailto:[email protected]>
>     http://mail.archive.org/cgi-bin/mailman/listinfo/ol-tech
>     To unsubscribe from this mailing list, send email to
>     [email protected] <mailto:[email protected]>
>
>

-- 
Karen Coyle
[email protected] http://kcoyle.net
ph: 1-510-540-7596
m: 1-510-435-8234
skype: kcoylenet
_______________________________________________
Ol-tech mailing list
[email protected]
http://mail.archive.org/cgi-bin/mailman/listinfo/ol-tech
To unsubscribe from this mailing list, send email to 
[email protected]

Reply via email to