Wikidata:WikiProject LD4 Wikidata Affinity Group/Affinity Group Calls/Meeting Notes/2019-07-02

From Wikidata
Jump to navigation Jump to search

Call details[edit]

Notes[edit]

Serials copyright project (John Mark Ockerbloom)

  • Collection of data about periodical copyrights trying to find information about historic serials that may not be under copyright anymore even though they were published after 1923
  • Might be out of copyright because
    • Until 1964, publications had to have copyrights renewed and in most cases serials did not renew copyrights or didn’t renew them from the start
  • Created inventory of all serials with copyright renewal available as part of MLS funded project
  • If there’s a serial of interest to your library, you could look it up in the list and see whether it had any renewals or not
  • More information link on the serial entry links over to Wikidata
  • Property in Wikidata Online Books Page publication ID
  • If you look up the periodical in Wikidata, you can link back to the copyright information and potentially actual online content
  • If there’s a Wikipedia article, it’s linked as well
  • Wikidata provides crosswalks to other data sources via links to identifiers
    • Means that if you have a subscription package and want to find out which have public domain issues or volumes we don’t know about, we can use the ISSNs, crosswalk it with a Wikidata ID and with these Online Books IDs to find out which things we know we have information about copyright or public domain from mid-20th century; can also find out which issues we don’t know about and potentially fill that in
    • We’re looking into ways of possibly doing that for serials and packages that are popular among libraries
  • Renewal data is stored as JSON files, which could also be included as Wikidata assertions
    • Not doing that because he’d have to define a mapping
    • Also saying renewals exist, but also these are all the renewals that exist--a little close world statement--possible to say that in an RDF style, but cumbersome to do that
    • Hard to say in Wikidata unless we define another property saying this set of renewals is in fact all the renewals that exist for a particular time period
      • If people know of ways to get around that, he’d be interested in doing that
  • Hilary: How are you going about adding links to Wikidata from the copyright project?
    • John: initially started entering data manually by searching for journals and other serials that exist; not all on list exist in Wikidata at this point; hasn’t started creating new ones
      • Looked into Mix ‘n Match tool, which seemed promising for doing a lot of links at once, but once you’ve uploaded the catalog of thing, you can’t update it yourself unless whoever writes the tool crawls your site and tries to update its own catalog, which seems cumbersome
      • Has SPARQL query to tell if data has been added to create links from my pages to Wikidata and Wikipedia
  • Forward to Libraries project
    • Can look up information about books by subjects
    • Makes connections between library catalogs
    • Has links to and from Wikipedia
    • Can also use correspondence between LCSH between Wikidata and Wikipedia identifiers
    • Insert articles from Wikipedia and link back to libraries
    • Has data on Github; some data is automatically calculated--you’ve got this LCSH and it corresponds to this Wikipedia article
      • Names are usually a one-to-one correspondence
      • Subjects can bring
      • A lot of these correspondences being put into Wikidata
        • Matt Miller found identifiers in LCSH and Wikidata and put in correspondences
        • Has not been done for not exactly the same matches
        • Interested in potentially using qualifiers
        • Can query Wikidata for correspondences and see if a better match has been added
        • Huda: The forward to libraries work seems very relevant to the discovery work we're doing in LD4P2 as well.  We may reach out to you about that later John if that's ok! (or if you're willing to speak about some of this later on a discovery call, that would be great too). Is this specifically to Wikipedia (and not necessarily Wikidata)?  Are these search string matches or matching names? Thank you for sharing!


Overview of Wikidata Policies slides

  • Questions:
    • JMO: Does the “highly used items” protection apply to both properties and “ordinary” entities?
      • Hilary: Looks like just items
    • Paul: I think a later discussion of the Notability policy might be a good contribution to LD4. I've heard the question "can/should I really add this as an item to Wikidata?" pretty frequently.
      • Hilary: Agreed. We’ll discuss notability in greater depth at our next meeting
    • JMO: FWIW, some properties in Wikidata can themselves be considered evidence of notability.  Recently, the property that links to my serial copyright information was designated as one such property.  So, I think that means that I (or anyone else) can create an item for a serial and have it considered notable in Wikidata as long as I have a copyright information page about it.
    • JR:    Would deletion of a duplicate entry typically require discussion, consultation of the community?
      • MaPo:    Not typically.
      • MePr:    No
      • Paul:    Not really.
      • MaPo:    For items, just go ahead and merge. If someone claimed two *properties* were playing the same role, that would require community discussion to sort out.
      • MePr:    But to be clear -- there is no discussion around merging, an editor does it at their own discretion
Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata:WikiProject_LD4_Wikidata_Affinity_Group/Affinity_Group_Calls/Meeting_Notes/2019-07-02&oldid=2119470155"