Wikidata:WikiProject University of Toronto Libraries/Discover Archives

Background edit

The goal of this initiative is to add "archives at" properties to Wikidata items for individuals or organizations that have corresponding public archival descriptions in UTL's Discover Archives database. Started in the fall of 2019, this initiative is in maintenance. Initially led by Kelli Babcock with significant support from Alex Wong and the many archivists at the University of Toronto. Our work began with the creation of a data model for the libraries and archives at UTL - see Wikidata:WikiProject University of Toronto Libraries/Data Model for Libraries and Archives. Following the creation of Wikidata items for archives at the University of Toronto, we used a csv export of top-level descriptions from Discover Archives to add "archives at" properties to Wikidata items. We used OpenRefine to complete this work. Our first batch of records added come from University of Toronto Music Library's Discover Archives 40-some holdings. More "archives at" properties were added throughout 2020-2021 and continues to be added as capacity allows.


The primary project goal of adding “archives at” statements to Wikidata is to make U of T archives more discoverable online. We want to link to the top-level description of each fonds/collection in Discover Archives. In archival description, the top-level description is what helps a researcher navigate all of the hierarchical information in a fonds/collection. We do not want to link to Discover Archives authority records. This means we are only adding “archives at” properties for Discover Archives archival records that represents a large portion of the person/organization’s archives - NOT lower level item or file archival records.


See Wikidata:WikiProject University of Toronto Libraries/Canadian Business Archives for work to explore sharing "archives at" properties with students and faculty.

Progress Report edit

Wikidata "archives at" items per institution can be retrieved from this query: https://w.wiki/62NQ

Institution Notes Items Created Query
University of Toronto Archives & Records Management Services Does not include "inventory number" qualifier See items
University of Toronto Music Library See items
University of Toronto Media Commons Already existing Wikidata items have "archives at" statements. Includes "Named As" qualifiers See items
Thomas Fisher Rare Book Library Some items have "archives at" properties with URLs to catalogue See items
University of Toronto Mississauga Library
University of Toronto Scarborough Library One already existing Wikidata item has "archives at" statement. See items
John M. Kelly Library Already existing Wikidata items have "archives at" statements. Newly created items for everyone but family fonds. 2 Family fonds still need items created. See items
OISE Library None of the Discover Archives holdings have already existing Wikidata items.
University of St. Michael's College Archives Already existing Wikidata items have "archives at" statements. Does not include inventory numbers because most do not have inventory numbers in Discover Archives See items
Trinity College Archives Already existing Wikidata items have "archives at" statements. See items
E.J. Pratt Library Already existing Wikidata items have "archives at" statements. Added VIAF/LoC IDs where possible. See items
Victoria University Archives Already existing Wikidata items have "archives at" statements. Added VIAF/LoC IDs where possible. See items


Process to link Wikidata to Discover Archives using the "archives at" property edit

  1. Explored options and workflows from other archives adding "archives at" property. Many thanks to archivists at York University's Clara Thomas Archives and Special Collections for their workflow and recommendations, as well as McGill University Archives presentations on this topic.
  2. Created a data model for adding archival and special collection institutions as Wikidata items (so that we could link to the specific archives when adding the "archives at" property). Data model found here: Wikidata:WikiProject_University_of_Toronto_Libraries/Data_Model_for_Libraries_and_Archives
  3. Exported data out of our Access to Memory (AtoM) database, Discover Archives, then edited and cleaned the data. Data exists across two tables, one for the top-level fonds and one for authority records. We want to link to the top-level fonds in Wikidata but require the metadata from authority records to create Wikidata items and populate them with basic statements.
  4. Communicated with and received input from the UTL Discover Archives Steering Committee on 1) what properties should be used along with the "archives at" property and 2) which Discover Archives holdings should have a Wikidata item and which should not (ex. controversial, going through changes).
  5. Reconciled the database export from UTL's Discover Archives AtoM database with Wikidata using OpenRefine and/or Google Sheets' Wiki plugin to identify which Discover Archives entities had Wikidata items already (thus only requiring an "archives at" statements to link Discover Archives to Wikidata) and which entities needed to be created from scratch.
  6. Batch added "archives at" statements through OpenRefine and/or QuickStatements for the entities that already had Wikidata items.
  7. Created items for people/organizations that did not already exist in Wikidata (for example - 9/42 items needed to be created when adding 42 fonds from UTL's Music Library archives).
  8. Added infoboxes into Wikipedia to link Wikidata/Discover Archives to the more popular Wikipedia interface to increase discoverability. [Still a work in progress!]

Data Model edit

Person - Basic Statements

Property Value Notes
Label Person's name Format: "First Name Middle Name Last Name", no dates or distinguishers needed
Description to help differentiate - ex. [Country] + [Occupation], "Canadian artist" Should be 2-12 words, starting with a lowercase letter unless it is a proper noun, no initial articles "an" or "a" or "the", and don't need periods. Used to differentiate and identify items - Description shows up in search results
Alias Alternative(s) to Person's name Variations including middle name, middle name initials, or no middle name. Alias will also be included in searches for item label.

Particularly useful for authority control; the most common name for a human may be different from the fonds’ name. Include the fonds’ name as an Alias.

instance of (P31) Instance of The class of the item. E.g. "human" (Q5) or "faculty" (Q180958)
date of birth (P569) Date of birth Add if recorded in authority record. Can be just year. Careful with PII (Personal Identifiable Information) for living persons. If using OpenRefine for importing, use format: YYYY or MMM, DD, YYYY. If using QuickStatements for importing, use format: +1967-00-00T00:00:00Z/9 or +1967-01-17T00:00:00Z/11 (See Help:QuickStatements for more details)
date of death (P570) Date of death Add if recorded in authority record. Can be just year. If using OpenRefine for importing, use format: YYYY or MMM, DD, YYYY. If using QuickStatements for importing, use format: +1967-00-00T00:00:00Z/9 or +1967-01-17T00:00:00Z/11 (See Help:QuickStatements for more details)
country of citizenship (P27) Country Add if known from the Actor's History. Value needs to be an existing Wikidata item. Property used retroactively, though there is discussion for humans born before the 20th century. e.g. Louis Riel existed before Canada became a nation, but he is considered a citizen of Canada. (See Property_talk:P27 for more details)
sex or gender (P21) Sex/gender Add if known from the Actor's History. While a fraught topic, it can be helpful in creating a SPARQL query to roughly know the ratios across collections. As gender is fluid, unless explicitly stated (ex. in the Actor's History) do not guess. Feel free to leave blank but if the biography includes a pronoun, this statement is used to understand the ratio of items of males to females and nonbinary genders, since the latter are historically underrepresented in Wikimedia. See the property notes for full list of gender options.
occupation (P106) Occupation Add if known from the Actor's History. Can focus on just the UofT related occupations. One human can have multiple occupations.
archives at (P485) Archives at Statement that is the impetus of these efforts and is specifically intended to describe the institution holding the subject’s archives. Include the reference and the qualifiers "inventory number" and "described at URL".
VIAF ID (P214) VIAF VIAF identifier may be included in the Discover Archives authority export.

Person - Optional additional statements

Property Value Notes
SNAC ARK ID (P3430) SNAC Ark ID
Library of Congress authority ID (P244) Library of Congress Authority Control
place of birth (P19) Place of Birth Can add if known
place of death (P20) Place of Death Add if known
cause of death (P509) Cause of Death Add if known
languages spoken, written or signed (P1412) Languages Spoken Written or Signed Add if known
educated at (P69) Education Add if known (universities or schools)
given name (P735) Given Name
family name (P734) Family Name (First)
employer (P108) Employer Add University of Toronto if employed there.
educated at (P69) Educated At


Organization - Basic

Property Value Notes
instance of (P31) Instance of Needs to be linked to a Wikidata item Ex. "academic department"
archives at (P485) Archives at Statement that is the impetus of these efforts. Include the reference and the qualifiers "inventory number" and "described at URL"

Organization - Optional additional statements

Property Value Notes
country (P17) Country Add value of country of origin e.g. Canada
located in the administrative territorial entity (P131) Located in the administrative territorial entity Add value of province of origin e.g. Toronto

Resources Consulted edit