RiC archival data available from the Swedish National Archives - feedback welcome!
253 views
Skip to first unread message
David Haskiya
Jul 10, 2023, 7:51:53 AM7/10/23
to Records_in_Contexts_users
Hi,
At the Swedish National Archives we very recently soft-launched a beta version of making our archival objects available as linked data objects in RiC and Schema.org representations.
Simple documentation in English here, RDF representation english version · Riksarkivet/dataplattform Wiki · GitHub
At the bottom of this page is the documentation for our RDF-mapping including a pragmatic choice we made in regards to how to map our volumes and dossiers to RiC. Feedback on that specific choice very welcome?
The linked data objects are intended to be used together with primarily our equally recently released Search-API, Search API · Riksarkivet/dataplattform Wiki · GitHub and our IIIF-APIs for accessing digitised archival records, IIIF english version · Riksarkivet/dataplattform Wiki · GitHub
As these are early beta versions we recommend the RDF-objects and Search API not be used in any production services, but we welcome all sorts of experimentation and feedback! Especially of course on the RiC-representation.
(Active development is on hold now for a while as we invite and evaluate feedback on this first public release.)
I should perhaps also make clear that this is an "out-data API" only. We have not made any changes at all to our internal data representations (either model or database structure or data). We have simply made it possible to on the fly, by dereferencing URIs, map the data in our (relational) database and (SolrCloud) search indices to RDF.
Best regards,
David
Magnus Sälgö
Jul 10, 2023, 10:59:06 AM7/10/23
to Records_in_C...@googlegroups.com
Very interesting thanks for the update…
Some questions before I tested it
1) linked data and internal structure
Do you plan to do “same as” any outside the National Archives?
I can see a community site like wikidata can be pointing at the National Archives as we do today but I feel much better is if you also point on Wikidata or on other sources…. My hope is that the historical newspaper database will move in direction Linked data and create “things” see request of an API et al in Swedish https://feedback.blogg.kb.se/forums/topic/lankmodell/
2) one user case I feel would be perfect for the National Archives is to implement the same functionality as DOI but for Swedish government documents that today is a chaos. Today it’s difficult to find documents and when the organisation is not longer alive and it’s archived at the Swedish National Archives it feels impossible….
Question: Are you looking into that right now or is it in your backlog? I tried to ask what other countries are doing but I feel this question is not on the agenda see https://groups.google.com/g/Records_in_Contexts_users/c/4zZy9MLz-Jg
3) OT: I see in the National Archives data that we have a lot of metadata debt… As I suggest to the historical Newspaper archive mentioned above they maybe should more involve volunteers and do add something like Wikidata with an UI and API. Any plans/discussion how to do the “internal” data better with a Wikidata like solution or?
--
You received this message because you are subscribed to the Google Groups "Records_in_Contexts_users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to Records_in_Context...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/Records_in_Contexts_users/47dc4ce4-c23a-45a1-97ce-bfc24bc262fbn%40googlegroups.com.
Richard Williamson
Aug 1, 2023, 12:13:05 PM8/1/23
to Records_in_C...@googlegroups.com, DHas...@gmail.com
Hi David,
Apologies for the slow reply! Thanks very much for this! I just wished to remark that your search API is quite similar to the one we have developed at the Norwegian National Archives (only available internally for the moment alas); I take this as a good design! In our case the actual data internally is represented in RiC-O and the search API design more or less directly reflects that, but it ends up being similar to yours.
If more institutions end up designing the same kind of API as this, it might be worth in the end trying to set up some kind of initiative to open-source an API which looks like this, to avoid every institution doing the same. One can envision that it might work in two ways: either one uses RiC-O internally, or one sets up a data mapping in some standardised way.
With regard to feedback on your mapping to RiC-O, I see nothing especially to remark upon; it agrees with how we see things. The only major difference is that we have not made use of the relation classes at all, but instead use things like rico:hasCreator directly. But it amounts to the same mapping conceptually.
Best wishes,
Richard
Apologies for the slow reply! Thanks very much for this! I just wished to remark that your search API is quite similar to the one we have developed at the Norwegian National Archives (only available internally for the moment alas); I take this as a good design! In our case the actual data internally is represented in RiC-O and the search API design more or less directly reflects that, but it ends up being similar to yours.
If more institutions end up designing the same kind of API as this, it might be worth in the end trying to set up some kind of initiative to open-source an API which looks like this, to avoid every institution doing the same. One can envision that it might work in two ways: either one uses RiC-O internally, or one sets up a data mapping in some standardised way.
With regard to feedback on your mapping to RiC-O, I see nothing especially to remark upon; it agrees with how we see things. The only major difference is that we have not made use of the relation classes at all, but instead use things like rico:hasCreator directly. But it amounts to the same mapping conceptually.
Best wishes,
Richard
--
Reply all
Reply to author
Forward
0 new messages