Skip to content

Wikidata Queries for Discovery

Steven Folsom edited this page Oct 27, 2020 · 5 revisions

Here we are collating follow on work from LD4P2 to understand connections between library data and Wikidata (https://github.com/LD4P/Wikidata/blob/master/Discovery/WikidataQueriesForDiscovery.txt)

As of 2020-10-27 Humans are the most likely entity type to be connected from traditional library authorities to Wikidata entities (991507). See: https://w.wiki/i$c

With this in mind, the initial focus here is on collating queries for useful information on Humans/Persons.

USE CASE 1: Persons associated with Cornell with an LCNAF, and lists of their notable and other works.

WHERE { SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],en". } ?CornellPerson ?p2 wd:Q49115 . ?CornellPerson wdt:P244 ?LCNAF . ?CornellPerson ?p1 wd:Q5. ?CornellPerson wdt:P800 ?NotableWorks . ?NotableWork wdt:P1476 ?title . } LIMIT 1000

Notes:

  • TODO figure out optionals.
  • TODO figure out different strategies for getting access to these works; this involve some combination of work identifiers and text searching the catalog, the bento box, and OCLC for ILL.
  • Test how much this actually exposes new works not already described in our existing discovery environment.
  • We can expand the query above to all persons by removing the association with Cornell (wd:Q49115).

USE CASE 2: Greater context for works, (for works with a history) inception and other dates related to works.

USE CASE 3: Greater context for works, Award winners. (pleasure reading, like texts, etc.)