MODS and RDF Call 2017-01-09
Time: 9am PDT / Noon EDT
Call-In Info: 712-775-7035 (Access Code: 960009)
Homework Reminder:
Moderator: Steven Anderson (Boston Public Library)
Primary Notetaker: Eben English (etherpad link: https://etherpad.wikimedia.org/p/RDF-MODS-20170109)
Attendees:
sanderson (BPL)
Eben English (BPL)
- Danny Pucci (BPL)
- Johanna Radding (Amherst)
- saverkamp (NYPL)
- christine.mayo (Deactivated) (Boston College)
- Betsy Post (Boston College)
- Juliet Hardesty (Indiana)
- mw2064 (Columbia)
- soriordan (Emory)
- Eric O'Hanlon
Agenda:
Conversion Code Update
- Nothing to report.
- Nothing to report.
Results of the Series and Collection voting.
- Spreadsheet: https://docs.google.com/spreadsheets/d/18JnrxIhIM5F17FWZ_i-JUckFDNHLENi0hRBMUJoNllc/edit?usp=sharing
- Respondents can see summary charts at: https://goo.gl/forms/Pl6a1ILMlWbMXLGV2 (doesn't show for others at the moment)
- 6 responses so far
- We will leave the survey open and try and solicit more responses
- Question 1: predicate mappings
- 3 votes for option 2: combination of dcterms:partOf (physical collection reference), opaque:memberOfSeries (series reference), pcdm:memberOf (digital collection)
- 2 votes for option 1: use pcdm:memberOf as the primary predicate for both series and collection
- 1 vote for option 3: same as option 2 except that it uses bibframe:hasSeries over opaque:memberOfSeries
- Question 2 & 3: Series and Collection Names/Titles
- 3 votes for full title mappings used for <mods:titleInfo>
- Would allow minting title objects for more complex titles
- Nobody has voiced any use cases that would require advanced functionality (nonSort, parallel titles, etc.)
- Would keep handling of titles consistent with what we have previously discussed
- 2 votes for rdfs:label
- 2 votes for dcterms:title
- Most people voting for "full title mapping" indicated they would simply use dcterms:title
- 1 vote for skos:prefLabel with optional rdfs:label
- Question 4: Series class
- Class needed to allow more consistency of data, facilitate querying by object type
- 2 votes for bibframelite:Series
- SirsiDynix may be implementing bibframelite as part of their work with Zepheira
- fact that this class doesn't exist in Bibframe 2.0 is a concern
- we should use established classes when possible
- 3 different Series classes in bibframelite – which one should be used?
- 2 votes for opaque:Series
- How would this be modeled?
- Subclass of pcdm:Collection? pcdm:Object?
- could add closeMatch to bibframelite Series class
- How would this be modeled?
- Archival Series vs. Publication Series
- We should use different predicates and classes for each
- Publication Series
- Already existing predicates in Bibframe for this
- Most people said they would not have a need for a publication Series class
- Archival Series
- Use mappings as agreed upon from voting and discussion
- Steven will attempt to create collaboration document with mapping for each
- We should use different predicates and classes for each
- OCLC identifiers in our identifier mapping
- See collab document at: https://docs.google.com/spreadsheets/d/1bJYgsX8b09dDvXjoiY87yk1hJ7KvAkLVYK_ycX-9700/edit?usp=sharing
- opaque:oclc could be used
- OCLC itself uses a local ontology which doesn't resolve (http://experimental.worldcat.org/ontology/library/oclcnum)
- dbpedia:oclc should be used (http://dbpedia.org/ontology/oclc)
- Other Collaboration Document discussion
- No further comments at this time
- We will need more discussion about classes for minted objects (titles, names, etc.)
Next meeting: Monday January 23rd at 9:00 AM PST / Noon EST