Time: 9am PDT / Noon EDT
Call-In Info: 712-775-7035 (Access Code: 960009)
Moderator: Steven Anderson (Boston Public Library)
Primary Notetaker: Shawn Averkamp (NYPL) (etherpad link: https://etherpad.wikimedia.org/p/RDF-MODS-20161128)
Steven Anderson (BPL)
- Shawn Averkamp (NYPL)
- Sara Rubinow (NYPL)
- Danny Pucci (BPL)
- Simon O'Riordan (Emory)
- Melanie Wacker (Columbia)
- Julie Hardesty (Indiana University)
Conversion Code Update
No updates for this week, but BPL will likely have updates to the code by next meeting.
Series, subseries, collection, institution Individual Institution Usage And RDF Voting Document review (https://docs.google.com/spreadsheets/d/19sjUN5oMsGHAVsBaZ-gncePuWPNPi7t5eOtxM5Sqml0/edit?usp=sharing)
3 options to put forth for voting
Where does bibframelite:Series come from? Not in bibframe 1.0 or 2.0. Seems to be only in bibframelite? Maybe a concern if we want to assert that as a type. Interesting that bibframelite has two different Series classes with two different URIs.
Using pcdm:memberOf may not work for institutions who cannot mint URIs or create objects for parent series/collections. It would have to be part of the migration to create objects for these host objects, which could be cumbersome for some.
Institution was not included in this mapping document because it seemed that maybe this was a use case specific to BPL. What is the consensus? What to do about external owing institutions? Wouldn't physical location work for this? Stephen will add institution as pcdm:Collection to the document, so people can vote on it.
Advantage is that you can get to collection or series more easily, but it seems that you would still need to resolve multiple levels of series to get to other subseries within the hierarchy.
Same as option 2, but uses bibframe:hasSeries instead of a custom predicate, but it is used incorrectly. The advantage is using a familiar predicate.
Option to represent collections/series as strings rather than objects is not present.
dcterms:isPartOf would require a non-literal value, would have to use dc:relation with literal.
Some institutions may not want to create new objects, so this would be an option. dc:relation, however, may not be the best mapping for this.
What to do about mapping items in a collection that do not include this information in relatedItem? This info should be in the rels-ext and you could use that to generate Collection objects.
Can IU post some of their examples where the 3 options posted wouldn't work? Julie will look into it.
Questions about bibframe series predicates. Why are we not comfortable using bibframe:hasSeries? Guidance isn't clear about semantics and use. Seems to be intended more for published series or serials, rather than archival series, so we really aren't using all of the series-related predicates correctly.
Stephen will put together a voting document once IU decides whether or not they can map their items to collection objects.
Other Collaboration Document discussion
No additional discussion items.
Next meeting: Monday December 12th at 9:00 AM PST / Noon EST