You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 4 Next »

Date: 

Attendees: 

Regrets: Jason

Agenda & Notes

Review actions from 2019-08-16 Cornell LD4P2 Meeting notes

  • Tim Worrall and Steven Folsom  to try the load tab with data from Discogs and an appropriate template.
    • Steven continuing to work on profiles; he's on vacation for 2 weeks starting 8/12 so expect no template progress.
    • 2019.08.16: 10 resource templates were pushed. Next week Tim Worrallwill have time to load those locally on his instance and then try to load n3 generated via QA. Update forthcoming
  • E. Lynette Rayle  will work with Michelle to develop a prioritization process. Steven Folsom is starting the slack thread b/t Michelle and Lynette
  • Simeon Warner to speak with Dave to better understand his capacity
    • Have discussed and Dave expects to catchup with backlog now request rate has eased a little and he is back to working full-time
    • Need to watch evolving prioritization process with Sinopia team to see whether that provides sufficient clarity
  • Simeon Warner meeting attendances
    • Steven, Lynette and Simeon to apply for LODLAM
    • Huda to attend BL summit
  •  Issues:

Status updates and planning

  • Kaleidoscope article: https://cornellprod.sharepoint.com/sites/cul/kscope/default.aspx#technical
  • Prep for Cataloging Sinatra and other 45's (Discogs data, https://github.com/ld4p/qa_server/issues?q=is%3Aissue+is%3Aopen+label%3ADiscogs)
    • ON HOLD pending more work in Sinopia to import data. Sinopia work through work cycle 4 will 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
    • Waiting for Work Cycle 2 to understand whether the derivation/cloning work item will happen during Fall
    • 2019.08.16: Sinopia 1.0 released and they're looking for bug testers. catalogers can start working soon. 
      • Tracey has a student working thru all of the 45s to identify which are not in discogs so we can catalog those first
      • Need remainder of resource templates completed and added to Sinopia. Then can start people working.
  • Enhanced Discovery (see also https://wiki.duraspace.org/x/sJI7Bg and https://github.com/LD4P/discovery/projects/1)

    • KPAOW plan: https://docs.google.com/document/d/1XuXH9n1YOhZY9cJhalA6ceTjOSpJrCsveoRgZyAUfwc/edit
    • Huda's user testing write-up
      • 2019.08.09: still being written. Will try and share when back from vacation
    • Linking Works to wikidata
      • 2019.08.16: Checked OCLC work ids in concordance file with wikidata... and about 1/600 return results. Worked thru about 5% of the concordance file. 
      • Can consider other use cases beyond clustering of works; examine which other data might yield results.
      • Series chronology, derivative works. Steven has some thoughts and will share with John.
    • Discogs within KPAOW 
      • 2019.08.16: demo from Tim - Two by Toot. WOAH! so much more data plus an image – along with highlights that reference data coming from discogs. 
      • Started working on linking genres to subjects... doing SOLR query on subject facet. 
      • Question: are you running checks to see if, e.g.: publisher is different between our catalog v. wikidata/discogs
      • Example of Coltrain at Newport '63 and Tyner live at Newport '63... need to ensure there are sufficient checks to prevent false matches
      • Application around directly identifying discogs URIs and/or 
    • Subject headings (demo by Huda)
      • knowledge panels for authorized subjects. at bottom of knowledge panel has digital collections results
      • FAST in JSTOR Forum for these items are not yet in SOLR. Using JSTOR Forum API. Will try to get a few examples working – to see if we have the connection, what it would look like.
      • next steps: working on navigation within the KP. 
      • link to digital collections and link to wikidata. When go to icon, how does user know which is which (UX concern). 
      • indentation increases screen real estate – interesting to address 
      • John has been working on the other influenced, expanding/collapsing, sorting.
  • Authority Lookups for Sinopia (Lookup infrastructure: https://github.com/LD4P/qa_server/projects/2, Authority requests: https://github.com/LD4P/qa_server/projects/1)
    • Dave loading all of the SHARE-VDE data to DAVÉ rather than focusing on only the institutions planning to use S-VDE data – 5 institutions have data available from SVDE, 1 (Frick) has data loaded in DAVE and now Lynette has to create config for these. Plan is to have an authority for each institution; use CKB to search across institutions (don't think DAVE has this data to load yet; maybe Stanford/Boulder/Alberta projects will rely on this)
      • E. Lynette Rayle will work on config for Frick early next week so that Stanford can test, also hope to get n3 export (below)
      • 2019.08.16: 5 new institutions being worked on. Alberta, Frick, Duke, Boulder, Cornell are complete. UCD, UCSD, Stanford, Yale are all in-progress. Dave should now have access to all institutions' data; process is time consuming to run.
    • Issue https://github.com/LD4P/qa_server/issues/162 is about getting n3 from QA to be imported into Sinopia (a different format from JSON or JSON-LD) - need to understand what data to get from SVDE and what profile to import into
      • Lynette planning to add something to QA UI to select authority, format and enter URI to do a fetch – will facilitate the copy-paste more easily.
      • 2019.08.16: Merged into dev but not yet into production
    • Lynette has created uber issue for LC authorities that are nearly there: https://github.com/LD4P/qa_server/issues/161 – want to get a number of these smaller issues done before dealing with the many new issues being created
      • 2019.08.09: from QA side, pushed all pending LOC work. need to confirm on cache side: extended context for all (Lynette needs to confirm all is coming back) AND genre subauth is active & deprecated. if search on deprecated, get active results so likely ignoring the subauth. In Sinopia and QA. Indexing issue. Dave has the action item here.
      • 2019.08.16: No movement yet
    • Hilary setting up meeting with Wikidata folks at Wikimania (Stockholm) around API and data questions documented by Lynette. Lynette will report if the API devs make changes to their output
      • 2019.08.16: meeting with wikidata dev team today. Put up basic search that uses their API (links shared via slack). Search is efficient but very limited data returned. Term fetch is super slow but returns beyond-everything. 
    • Currently prioritizing LC and SVDE authorities pending further input from Michelle re cohort priorities
      • will be working with michelle and steven on getting this prioritized. Michelle gave some high-level priorities that align with our current work but the mass of requests are not yet prioritized / there is not yet a process for this
    • Boosted performance. DEMO!!!!: performance in graphs - 24 hours, 30 days and 12 months. Started running this on 8/15. A few authorities must be consistently doing worse than others... avg for all requests is just shy of 2 seconds. Browsing thru log, most are sub-second...
      • Thru-put testing has not been set-up yet. Theoretically on elastic beanstock so should adapt with limited concern to higher hit-rates
      • Next step: subset more statistics to see whether there are authorities performing consistently worse than others. Could just be amount / quantity of data being returned. Also want to subset by term fetch versus searches. 
    • New addition: can do a term fetch in the UI. Does not yet do this for discogs - should happen. Can request in json, json-ld, n3 (needed to paste into Sinopia) via QA server. Can do a config for discogs
      • For Sinopia, need to specify resource template in the data for load rdf tab. Shouldn't be within QA itself, anyway... since so Sinopia specific
  • Travel and meetings (see LD4P2 Cornell Meeting Attendances)
    • LD4 BL meeting September 23 week in Stanford
      • Huda, John, Steven going
    • European BIBFRAME Summit in September 16-17th-ish
      • Jason going and ARM/rare-cohort proposal accepted
    • Blacklight Summit will be at Duke, 9, 10, 11 October at Duke
      • Huda
    • Samvera Connect, week of October 21 (WUStL)
      • Lynette to present on QA
    • Fall partner and cohort meeting in DC, November 12/13
      • Everyone should plan on attending
    • 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 2020
    • rdfs:seeAlso Conferences Related to Linked Data in Libraries
  • Next meetings:
    • Jason out Aug 30
  • No labels