On 11 July 2014 21:15, Eric Lease Morgan <eric_morgan@infomotions.com> wrote:
The ultimate goal is to discover the feasibility of exposing bibliographic and authority content in YAF — “linked data”.
Awesome! This might be redundant, but I'd like to point out there are some efforts in this direction already: Some thougts from me are on the wiki: http://wiki.koha-community.org/wiki/Linked_Data_RFC Chris Cormack has been working on an automatic conversion from MARC to RDF-in-a-triplestore when records are saved in Koha: https://gitorious.org/koha/mainline/commits/ff19b7040d550811ec0f5eaee93f6fb6... (Not sure if that is the best link for it) (I think Chris would like to rewrite this to use Catmandu::Exporter::RDF) I have been working off and on on a Linked Data Browser for Koha: http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=10788 By far the most ambitious project related to Koha and RDF/Linked Data (that I know of) is the one at Oslo Public Library (where I am currently on paternity leave from a part time, temporary position). Their aim is to have RDF as the master data format (which includes cataloguing in RDF) and push MARC-records converted from RDF into Koha, so Koha can do circulation and other menial tasks. ;-) Some relevant blogposts: http://digital.deichman.no/blog/2014/01/16/oslo-public-library-chooses-the-f... http://digital.deichman.no/blog/2014/06/19/oslo-public-library-chooses-rdf-l... http://digital.deichman.no/blog/2014/07/06/rdf-linked-data-cataloguing-at-os... They/we have stuff like a marc2rdf converter: https://github.com/digibib/marc2rdf and are working on a cataloguing interface (independent of Koha): https://github.com/digibib/armillaria
If all Koha implementations were to make their content available as linked data, then the whole would so much greater than the sum of its parts.
For me, just converting MARC to RDF is not that interesting or useful. It gets interesting when - we start to have proper identifiers for things, instead of text strings, and when - librarians can start to augment the data from the MARC records with new kinds of data (that would never have fitted into MARC) that we can then use to create useful new features Let's get the conversation going! :-) Best regards, Magnus