Date:
Attendees: Lynette (zoom), Tim, Steven, Huda, Adam, John, Simeon
Regrets: Jason
Agenda & Notes
Review actions from 2020-01-10 Cornell LD4P2 Meeting notes
- Huda Khan to discuss with Astrid and David possible collaboration with U Chicago over usability (and maybe others in DOG team)
- Notes: https://docs.google.com/document/d/1_V7JfAKqSn63-G0zupLZigt8o7R3v_WXC7hWlt8w744/edit
- 2020-01-10 Chicago report will be discussed in DAG meeting on Jan 21 (expect to happen as scheduled)
- E. Lynette Rayle QA performance
- Dave made some changes to address issues with 50x responses, still possible issues under high load (possibly something on the Sinopia side but waiting to look at IP address to logging to identify whether same source is hitting us with same request)
- Ongoing discussion about the need build more complex indexes to deal with slowness of complex queries
- 2020-01-17 - no recent changes, Dave is out
- See items under travel/conferences
Status updates and planning
- John has attended to D&A meetings with user reps, showed video searching Google Books to deal with the zero results case which they found interesting. Adam notes that much D&A work is quite low-level and we don't have a good way to think about the big ideas. Could we have a server for labs/beta?
- John had discussion with HT about API which is on their to-do but not scheduled. May be open to providing index access
- Cataloging Sinatra and other 45's (Discogs data, https://github.com/ld4p/qa_server/issues?q=is%3Aissue+is%3Aopen+label%3ADiscogs)
- Usability - suffering from long vertical display with expanding sub-forms, lots of scrolling
- Discogs QA module input is only using some data based on the profile being used, trouble filling in only work and not instance data when working from a work form with embedded instance. Steven will investigate whether template or profile changes can address this – 2020-01-17 Steven did some work to see whether filling in of multiple forms work differently depending on where you start. Found that starting with master doesn't add instance, starting with instance does add master too. Tim reports that this is intentional in the design of the Discogs authority, one could use the Discogs notion of "main release" to add one release when a master is loaded but this isn't done and might not be right ==> conclusion is that cataloging should start with a release/instance
- Lookup performance still an issue but not even or even consistent within the same lookup. Seems also that there is an effect from overall load on the service. → Steven/Lynette/Huda to work out where the problem lieslies – 2020-01-17 being scheduled next week (as I write)
- Lookups for place not usable, relies on work from Dave to fix: https://github.com/LD4P/qa_server/issues/248 & https://github.com/LD4P/qa_server/issues/240
- Work is being done but places are not being recorded
- Enhanced Discovery (see also https://wiki.duraspace.org/x/sJI7Bg and https://github.com/LD4P/discovery/projects/1)
- BAM! (to run through end October)
- For reference: Initial draft of BAM plan
- BAM Screenshots (Tim's call number screenshots added.)
- Finishing up lessons learned
- SMASH! (to run through end of January)
- John showed demo of search text in Google Books for cases where we have no hits in the catalog, using results to find items in the catalog via ISBN. Want to expand to use OCLC number.
- Huda has LCSH lookup feeding into query of BAM! index to get birth and death dates, and then also getting other people born within that lifespan in order to show contemporaries; also working on influence relationships – question of how to get sufficiently closely related entities
- Huda has been looking at annif – will train using text from our index and assigned subject headings to learn the mapping which might then be used for lookup
- John looking at the cases of queries that return small numbers of bad search results. Should "a secret history" return "the secret history", especially if there a few matches from the default search? John will continue to explore...Huda showed demos doing string matching (non left anchored) in LSCH and then finding broader and narrower subjects. Going to do expanded search of broader and narrower for the facets that come back from the catalog text search.
- Open meeting March 3, 2-3:30pm in Mann 102 and should Zoom it too
- How will we decide what to take forward from KAPOW!, BAM! and SMASH!?
- BAM! (to run through end October)
- Authority Lookups for Sinopia (Lookup infrastructure: https://github.com/LD4P/qa_server/projects/2, Authority requests: https://github.com/LD4P/qa_server/projects/1)
- See above
- Monitoring tests have been adjusted to run at night, required work with Robbie to adjust pingdom
- Travel and meetings (see LD4P2 Cornell Meeting Attendances)
- 5th International LODLAM SUMMIT at the The Getty Center in Los Angeles. February 3-4, 2020
- Steven, Lynette and Simeon will attend
- E. Lynette Rayle will put QA in for the "tool challenge": CFP due 1/17
- code4lib - March 8-11, 2020. proposals done, registration still open
- Knowledge Graph Conference (Columbia University), May 6-7, workshops 5/4-5/5
- Workshops due 1/15 (not doing this), proposals due 2/28, talks 20mins
- Huda Khan will lead effort to consider proposal, discuss next week 2020-01-17
- LD4 Conference at College Station, TX (TAMU) - May 13/14, 2020
- Proposals due 2020-01-31 – https://stanforduniversity.qualtrics.com/jfe/form/SV_ebmQF48Gn9bag6x
- Lynette will consider lookup best practices and discuss with OCLC
- Steven/Tim to think about discogs "supervised conversion"
- Huda/John/Tim/Steven to think about KPAOW! BAM! SMASH!
- Simeon Warner to check on travel budget
- rdfs:seeAlso Conferences Related to Linked Data in Libraries
- 5th International LODLAM SUMMIT at the The Getty Center in Los Angeles. February 3-4, 2020
- Next meetings: