E. Lynette Rayle will describe current state of wikidata work (as identified on Prioritization page) and then ask Christine Fernsebner Eslao what we might consider "done" for the next pass on wikidata, then perhaps create new and more specific issues for later improvements
2019-10-04 Not done yet, still need to check in with Christine (have been working on other items higher in prioritization)
Tim Worrall to update discogs code to use standard bnode format
2019-09-13 Tim about half way through this
2019-09-27 Have updated discogs code to output ntriples as well as n3, will make PR for this and also the update to use bnodes
2019-10-04 Got back from Justin a snippet of n3 that now works with Sinopia, and have now replicated this load of n3. Sinopia team are updating the label to list all formats that can be accepted. Have made PR which is awaiting review from E. Lynette Rayle. Have found a new issue with loading Discogs data in that certain extra labels and types break the loading of data → need to identify minimal example(s) to post bug report for Sinopia and potentially use for a workaround. This seems like something that is a general issue and will be a problem with other data sources such as SVDE. General idea is that extra triples should be ignored or loaded without showing, should not cause load to fail
Tim Worrallwill create issue to describe that Sinopia doesn't accept n3, it accepts only ntriples → see above, that has changed
Awesome organization by Jessie and Huda, attendees needed lots of support and facilitation after 2.5 days (it was a pretty intense meeting and people did a great job participating), only a few people stayed for the half of the 4th day
Jessie, Mark, Chris (Stanford) and Huda are going to BL meeting next week and will take ideas forward
Succeeded in synthesizing scenarios with a good mix of BL and other folks (e.g. Vufind, EBSCO, Ghent autocomplete, LC)
Specific suggestion to implement schema.org beyond microdata to have a JSON-LD serialization as focus of BL output
AWAITING more work in Sinopia to import data. Sinopia work cycle 2 will we hope include the ability to read in RDF back from Trellis. We hope that we can leverage this to import RDF from a lookup in Discogs or ShareVDE. STARTING with items not in DIscogs
Had kick-off meeting on Wednesday (Oct 2) and will follow up with a review of BIBFRAME and how it will be used for this work, and training around the tooling. Catalogers has accounts with Sinopia etc.. Steven adding Sinatra training material to the Sinopia drive space, immediate feedback is that some understanding of the model is required to be able to think about things like work vs instance title
Finding issues with sub-entities being tied to profile for the parent limiting editing flexibility (e.g. instance is tied to the particular work profile)
Did some evaluations on mockups at Stanford last week with freshmen undergrads, some of whom had never yet used the library catalog. Included broader and narrower in subject facets; and offering different browse axes for author, subjects, geographical and timeline
In item view, Tim looking at call number browse and expanding work on subject browse started in K'PAOW!
For timeline, John looking at subject era facet that depends on FAST which has labels and connections
Huda looking at querying subject headings, discussed with Dave and running long queries to extract data set. Looking at fairly coarse grained set of top level subject headings as a possible entry point
For geographical browse looking at subject areas for regions. In our current catalog the data is quite sparse
RDA registries included as a single authorities with sub-authorities for the different areas, now available in Sinopia. This has been straightforward because of consistent SKOS model and clean data
Open question about cohort members who want to use LC versions of RDA vocabs. Discussion on id.loc.gov email list and Matt Miller asking for feedback on data dumps
Performance – analysis on laptop clearly shows that from QA server perspective the http load is the dominant part of the problem with the slowest third of requests taking 2.5-6s for load. The normalization part is next slowest, taking ~0.8s for the slowest third. Need to check analysis is correct and also run on AWS. Will discuss the load portion with Dave. And also do swarm test.
Tim will be on D&A sprint that week, everyone else should plan on attending
Team bus will leave Ithaca ~2pm Monday, return Thursday morning. We'll stay around Dupont, perhaps Holiday Inn or Darcy next door? Simeon will get the bus
5th International LODLAM SUMMIT at the The Getty Center in Los Angeles. February 3-4, 2020
Steven is on the planning committee, Lynette and Simeon to apply
Expect to have a "tool challenge" - a competition before the conference
LD4 Conference at College Station, TX (TAMU) - May 13/14, 2020