Wikidata:Tools/inteGraality

inteGraality is a tool for querying Wikidata by creating a dashboard of property values for a specific set of items.

The dashboard is configured using the {{Property dashboard}} template, and a bot will then update it − every week or on demand.

When inspected, the display of the queried Wikidata creates room for more data-driven questions that, when further analyzed, would provide valuable insights into the questioned areas and reveal data inaccuracies.

• Select a subset of Wikidata

  all album (Q482994), all red panda (Q41960), all painting (Q3305213), all British MPs…

• slice it along a dimension (ie, a property)

  red pandas by residence (P551), paintings by collection (P195), albums by genre, language, year…, camera models by manufacturer (P176)

  • optionally group the slices along a higher dimension

    the country of the collections of the paintings, the country of the residence of the red pandas, the continent of the country of the lighthouses

• choose how the subset should be described: properties, labels, descriptions

• Surface errors of grouping

  5 albums published in 1849

• Identify patterns

  Most Russian rock (Q2624785) albums are missing a producer (P162) ; all paintings in the Condé Museum (Q1236032) collection have an image (P18)…

• Work on completion piece-meal − a given property for a given grouping

  Work only on the Rijksmuseum paintings missing a title (P1476), on the Dreamcast video games missing a genre, etc.

The selector_sparql can be used in the most straightforward way (like -instance of (P31)painting (Q3305213)), but you can also make more advanced selection. Some examples:

• Humans with National Library of Wales Authority ID (P2966): |selector_sparql=wdt:P9495 []; wdt:P31 wd:Q5
• Alumnis of two related universities: |selector_sparql=wdt:P69 ?edu . VALUES ?edu { wd:Q309350 wd:Q6806876 } .
• British MPs: |selector_sparql=p:P39 [ ps:P39 [ wdt:P279 wd:Q18018860 ] ]
• Collection highlights National Library of the Netherlands: |selector_sparql= wdt:P195 wd:Q1526131; p:P195 [ ps:P195 ?id ; pq:P2868 wd:Q29188408 ]
• Editions with a sitelink to French Wikisource: |selector_sparql=wdt:P31 wd:Q3331189. ?sitelink schema:about ?entity; schema:isPartOf <https://fr.wikisource.org/>

You can (ab)use the higher-grouping mechanism to display the flag image (P41) of a country-grouping: |higher_grouping=wdt:P41 (example)

On the face of it, the grouping parameter just takes the name of a property, eg `P475`. However as this is used to construct SPARQL queries, you can inject additional SPARQL to achieve creative things.

Example: User:Multichill/Windmill_sandbox − using |grouping_property=wdt:P131/wdt:P131 allows to group the windmills one administrative entity above.

Example: WikiProject Music/Statistics/Albums by country of performer − the performer (P175) of an album can be a person (the country is then country of citizenship (P27)) or a musical group (country is country of origin (P495))

You can inject the | SPARQL “OR” operator − but it must be escaped with {{!}}: |grouping_property=P175/(wdt:P495{{!}}wdt:P27)

You can limit the domain of the groupings by adding a SPARQL condition − using a reverse path expression so that the final query is still valid: |grouping_property=P39 ?grouping . wd:Q16707842 ^wdt:P279

Let’s take the example of member of the Chamber of Deputies of the Italian Republic, whose office is modeled as:

You can group per parliamentary term (P2937) using this expression: |grouping_property=P39 wd:Q18558478 ; p:P39 ?pstmt. ?pstmt ps:P39 wd:Q18558478 ; pq:P2937

The properties (aka columns) parameter has support not only for properties, but also qualifier:

• P39/Q18558478/P580 will check if the relevant position held (P39) statement has a start time (P580) qualifier − -position held (P39)member of the Chamber of Deputies of the Italian Republic (Q18558478)start time (P580)xyz (example)

• wikitech:Tool:InteGraality
• The Wikipedia Workbook for Cultural Institutions (2024) has a user guide for inteGraality starting on page 42. (see phab:T376902)