MODS and RDF Call: 2018-10-29

Owned by Eben English
Last updated: Nov 16, 2018
3 min read

Time: 9 AM PDT / Noon EDT

Call-In Info:  712-775-7035 (Access Code: 960009)

Documents/Homework:

Moderator: Eben English

Notetaker: TK (Etherpad: https://etherpad.wikimedia.org/p/MODS_and_RDF_Call__2018-10-29)

Attendees:

  • Melanie Wacker (Columbia)
  • Julie Hardesty (Indiana)
  • Emily Porter (Emory)
  • Eben English (BPL)

Agenda (with notes)

  1. Feedback Form Responses (no new entries as of 10/29)
    1. https://docs.google.com/spreadsheets/d/1DRAzr3ExEAjK2RUs63gxcymoDwprQjVRPDsU_3dDGVU/edit?usp=sharing

  2. Email Responses (no new entries as of 10/29)
    1. https://docs.google.com/spreadsheets/d/146uvvKt5ZVvWjm9WV28fWHgCP8fYQ_JXkc5GcMv98Wc/edit?usp=sharing

  3. Samvera Connect Session: Slides

    1. small audience but interested

    2. question about process for requesting minting predicates

    3. mapping and scope is complex so implementing might seem daunting to community

    4. maybe ask for stories to be shared at next Samvera Connect? (Avalon, Northwestern)

  4. Action items from last meeting:
    1. Add comments to recordInfo direct mapping  DONE

      1. leave direct mapping as it is with comment about technical problem

      2. minted object updated to show minted metadata object

    2. Update recordInfo minted object mapping to use minted metadata object  DONE
    3. Add bf:variantType to <mods:title> minted examples  DONE
      1. bibframe variant added to MODS title for uniform, etc titles
    4. Remove instances of skos:exactMatch and skos:closeMatch where appropriate; replace with owl:sameAs and rdfs:seeAlso  DONE
    5. Inquire about bibliotek-o status DONE
      1. bibliotek-o status checked - project is dead and shouldn't be used

  5. Outstanding Comments on Recommendations document
    1. <mods:title> (minted object)
      1. "supplied" designation
        1. bibliotek-o is deprecated, we should not use these
        2. BF has no support for this concept
        3. Could simply recommend that supplied status  be indicated by brackets in title value and/or explanatory note
        4. Could use opaquenamespace
        5. can't recommend to use brackets because RDA guidelines so there isn't really a solution; recommend to follow content standards being used - DACS and RDA have different recommendations
        6. Eben removing bibliotek-o and add note at beginning of section regarding content standard; use skos:note to indicate in Example 5 (p. 66)
      2. bf:Title class used but not bf properties within that class
        1. helps with consistency across recommendations to use same properties (across simple and complex); also more generally usable beyond BF-centered cases; implementations can make use of BF properteis within Title class if desired but showing that it's not necessary


    2. MODS note
      1. having noteType as subclass helps with interoperability (everyone using the same note type) but BF not creating controlled vocab so there are no type URIs to use here
      2. simpler to keep note types as text here
      3. keep example as is with text and add info about best practice is to create subclass for note type
    3. mods subject
      1. pre-coordinated subject headings questions
        1. we approached this is a FAST-ish way to explode pre-coordinated subjects but could be problematic if there are multiple pre-coord subjects
          1. FAST philosophy is that pre-coord isn't helpful to begin with anyway
        2. Julie will try Example 7 using MADS/RDF and see what happens
        3. FAST vs LCSH - how comparable are they? should be FAST equivalents except for when there are free-form combinations in LCSH that aren't as easily automated
        4. FAST approach is faceted application applying subject pieces as you need them without worrying about order and combination
        5. also need better documentation at beginning of this section
      2. geographic coordinates as subject p. 95 example 12
        1. comment supplies really complex recommendation that requires even more object minting
        2. DCMI box and point specs are getting old (2006) but not sure if they are out of date or not
        3. Samvera Geopredicates WG documentation: https://drive.google.com/drive/folders/0B3s6m1gxi006N3p1anBpbmZuOEU
        4. might be schema.org predicates that can help (schema:latitude, schema:longitude, schema:box) - might be more accessible
          1. expected to be used with coordinate and space objects instead of place objects so that might be problem or complication
        5. could stick with edm:Place class only; could use rdau class for geographic object: http://rdaregistry.info/Elements/u/P60109
        6. examples work now, information is all there, machine readable if you know content standard being used, programming logic needed to use info
        7. maybe keep example as is and add information at beginning that there are other options for providing coordinate-based information depending on needs

  6. TODOs:
    1. Submit opaque predicates to URI WG

  7. Next meeting: Monday November 26