JSON-LD context and framing for BIBFRAME data in the Bluecore Project

If there’s time I’d like to talk briefly about our use of JSON-LD context and framing for BIBFRAME data in the Bluecore Project, and make a case for shared community practices around BIBFRAME JSON-LD data.

12 Likes

Good use of that Drake meme! :slight_smile:

1 Like

@edsu For a SWIB18 workshop we transformed the Bibframe works dataset into LOUD and wrote about it in this blogpost: Indexing the Bibframe works dataset as JSON-LD | lobid.org At the time, there no canonical JSON-LD context for Bibframe existed. Is there one available now?

1 Like

Thanks for sharing this here @acka47! As far as I know there is not a canonical JSON-LD context for Bibframe. Maybe it is at odds with the needs of local organizations? But I think it is a serious impediment to interoperability by people who are not already invested in RDF tooling. Perhaps it’s worth revisiting?

I just looked a bit more into it and saw that I had a short discussion about the context with Kirk Hess, see Collecting feedback on LoC dataset · Issue #33 · hbz/swib18-workshop · GitHub . If LoC is not working with JSON-LD themselves, I guess they will probably still not see the need for an elaborate @context document.

@acka47 Yes, I think that’s right. Thanks for the link to that conversation. I believe LC’s bibliographic systems are still XML-centric, which is fine in itself. But I think if the BIBFRAME vocabulary is going to be widely used, effort must go into arriving at a way to also share it as JSON? It might be the case that this work has to happen around LC if they don’t see the immediate value?

1 Like

I wrote a brief blog post with a recording of me practicing the lightning talk, and some notes from this conversation: LOUD BIBFRAME?

Thanks for the post! With regard to the SWIB18 workshop it says, " there was a SWIB18 Workshop focused on creating a JSON-LD Context for BIBFRAME data. Actually, the workshop was about creating LOUD from RDF in an intuitive structure and so that it can be queried in complex ways, about indexing and using the data. It also included framing (though a very simple frame sufficed), see the slides/overview at From LOD to LOUD: making data usable. The context was just a by-product we created beforehand. Basically, we applied all the lessons learned from publishing LOUD in lobid and transferred them to the Bibframe example work dataset published by LoC at the time. See also the repo for the workshop.

[Edit: My colleague Fabian reminded me that we did a second rendition of the workshop at ELAG 2019 which probably was a bit more polished: slides, repo.]

1 Like

@acka47 thanks, I hadn’t followed the link yet, so I appreciate the clarification. That was quite a thorough workshop! Given you were trying this 7 years ago, maybe it’s not worth it to keep beating this drum? But it was a good idea then, and I think it is still a good idea today, especially when you look at the success of the fediverse & iiif.

I definitely think, you should keep beating this drum. As you say, it works very well for ActivityPub/Fedi and IIIF. We at hbz are still very happy with our LOUD approach for lobid, especially in terms of the maintenance effort to usefulness ratio. (I want to say: We have been running for several years a highly available, very usable and in fact heavily used LOUD service/JSON-LD API with relatively few personal and technical resources. This would not have been possible based on RDF.)

1 Like

In the conference chat @niklasl commented:

Thanks for the talk! At NatLib Sweden we publish (and store+manage!) Libris, the union catalogue, as compact JSON-LD (among other formats; see links in Lawrence in Arabia · war, deceit, imperial folly and the making of the modern Middle East. · Atlantic Books · 2014 | id.kb.se or content-negotiate on that). Also see our docs: Read a record - Libris API documentation

@edsu feel free to use this forum for continued discussion on the topic. We will soon remove the navigation elements for the conference (banner, dedicated navigation on the left) and move back to a pure forum structure. Generally, we invite everybody to utilize the forum to bring discussions on standards, implementation questions etc. to the SWIB community, also and especially in between the conferences.