tutorial

Date added: 17.10.25

How to query Wikidata

A practical guide to querying Wikidata, through three different endpoints. The guide breaks down a complex query to provide insights into how Wikidata is stored and structured.

Contributor:
- Sotirios Alpanis->
Title:
How to query Wikidata
Last updated:
17.10.25
Description:
A guide to working with Wikidata
Tags:
- Wikimedia
- Python
Experiment:
Connecting collections with Wikimedia->

CC BY 4.0

Launch Colab Tutorial->

How to query Wikidata

Wikidata is a free, collaborative knowledge base that stores structured data about almost everything. It is the database behind Wikipedia, and as such, its sheer size and potential can be a little overwhelming at first. This is a hands-on guide to help get over that initial barrier.

During Wikifest 2025 Wikidata expert Alex Lum was helping with research data about female authors, specifically during the lifespan of Jane Austen. Alex created and shared this SPARQL query:

SELECT DISTINCT ?author ?authorLabel ?genderLabel ?countryLabel
WHERE {
  wd:Q36322 wdt:P569 ?ja_birth .
  wd:Q36322 wdt:P570 ?ja_death .
  ?publication wdt:P31 wd:Q47461344 .
  ?publication wdt:P50 ?author .
  ?publication wdt:P577 ?pubdate .
  FILTER(?pubdate > ?ja_birth) .
  FILTER(?pubdate < ?ja_death) .
  ?author wdt:P27 ?country .
  ?author wdt:P21 ?gender
  SERVICE wikibase:label { bd:serviceParam wikibase:language "en" }
  }

The query, and its results can be seen on the Wikidata Query Service https://w.wiki/FNSu. However, for those unfamiliar with Wikidata and/or SPARQL, simply seeing the results may not help understand what's going on. This guide will use the query as the basis for an introduction to Wikidata.

What is SPARQL?

SPARQL is a language used to query graph databases, specifically the Resource Description Framework (RDF). Graph databases differ from more traditional databases because information is stored as connections between things, like a web, network or graph.

Note: SPARQL is a recursive algorithm - 'SPARQL Protocol and RDF Query Language' - and is pronounced sparkle.

Triples format

In graph databases data objects are stored as 'triples' - statements with three parts:

Subject -> Predicate -> Object
e.g. "Paris" -> "is capital of" -> "France"

A database comprised of triples builds up into a web, or network of linked data. Which SPARQL lets you query. For example if we add the following to the triple above:

"Paris" -> "has population of" -> "2,100,000"
"London" -> "is capital of" -> "United Kingdom"

Then we can use SPARQL to retrieve information about Paris, but also about capital cities.

Note: the triple approach to describing data and its relationships results in it being referred to as 'semantic'.

Introducing SPARQL (and Wikidata) syntax

If you are familiar with other database query languages, like SQL, then the overall structure of Alex's SPARQL query will have looked familiar. Even for those unfamiliar with database query languages, keywords such as `SELECT`, `WHERE` and `FILTER` are all descriptive and give a sense of what's going on. However, there are a few less obvious bits of syntax:

?<variable_name> is used to declare a variable, i.e. a reference where a value can be stored. In Alex's query ?author is a variable in which any authors that are retrieved are stored.
wd - Wikidata entities, i.e. items and properties that are being retrieved from Wikidata by their id.
wdt - Wikidata 'truthy' statements, i.e. property values. Gives access to the preferred (or "best") property values for an entity. Wikidata can store multiple values for the same property, which can be marked as preferred, normal or deprecated. `wdt` will prioritise returning the preferred values, then any normal values where preferred is not present.

e.g.

SELECT ?capital WHERE {
  wd:Q142 wdt:P36 ?capital .  # Q142 = France, P36 = capital
}
# Returns: wd:Q90 (Paris)

Querying Wikidata

Wikidata has a few different ways of querying its data. AlthoughAlex's query is written in SPARQL, we'll use it as the basis for exploring the following endpoints:

Search
Linked Data Interface
Wikidata Query Service (SPARQL)

Authorising requests to Wikimedia

Although querying Wikimedia doesn't require you to authenticate via e.g. an API key, it does require users to create a User-Agent header to help manage fair use.

Note: Wikimedia documentation related to setting a `User-Agent` header: https://foundation.wikimedia.org/wiki/Policy:Wikimedia_Foundation_User-Agent_Policy

We can use the same authorisation header to query each of the three data endpoints:

wiki_auth_headers = {'User-Agent' : 'SLV-LAB-Bot/0.0 (https://lab.slv.vic.gov.au/; [email protected])',}

Search

The simplest Wikidata endpoint 'Search' allows you to query using a search term. The search is fuzzy, so results will include things that match and partially match the search term. This makes it useful at the beginning of research where the specific identifier for a Wikidata entity may not be known.

Note: Wikidata provides a web interface for performing search: https://www.wikidata.org/wiki/Special:Search

The basis for Alex's search was Jane Austen, so here's how the Search service can be used to retrieve data related to the search term 'Jane Austen':

import requests

search_term = "Jane Austen"


url = "https://www.wikidata.org/w/api.php"
params = {
    'action': 'wbsearchentities',
    'format': 'json',
    'language': 'en',
    'search': search_term
    }
    
response = requests.get(url, headers=wiki_auth_headers, params=params)
data = response.json()
results =  data['search']

for item in results:
    print(f"ID: {item['id']}, Label: {item['label']}, Description: {item.get('description', 'N/A')}")

ID: Q36322, Label: Jane Austen, Description: English novelist (1775–1817)
ID: Q96085049, Label: Jane Austen, Description: (1655-1717) wife of Stephen Stringer
ID: Q75494834, Label: Jane Cranley, Description: Peerage person ID=176244
ID: Q75631099, Label: Jane Austen, Description: (born 1826)
ID: Q54639704, Label: Jane Austen, Description: 1986 book by Tony Tanner
ID: Q20669108, Label: Jane Austen, Description: Russian encyclopedic article
ID: Q339693, Label: Mafia!, Description: 1998 film by Jim Abrahams

Linked data interface

Provides links directly to individual entities or properties, but requires the Wikidata ID to be known.

In our example, now that the Wikidata ID for 'Jane Austen' is known Q36322, it can simply be added as a parameter to the URL http://www.wikidata.org/entity/Q36322

The URL can be modified to retrieve different information:

/entity/ gets the entity
/Special:EntityData/ retrieves data about the entity

.json can be appended to the URL to return the data in the a machine friendly JSON format.

wikidata_id = 'Q36322'

linked_data_uri = f"https://www.wikidata.org/wiki/Special:EntityData/{wikidata_id}.json"

response = requests.get(linked_data_uri, headers=wiki_auth_headers)
data = response.json()

entities = data.get('entities',{}).get(wikidata_id)

labels = entities['labels']
label = labels.get('en').get('value')

descriptions = entities['descriptions']
description = descriptions['en']["value"]

print(f"{wikidata_id}: {label} - {description}")

Q36322: Jane Austen - English novelist (1775–1817)

Now that we have a way of retrieving data for one Wikidata ID, let's pull out each of the Q and P IDs from Alex's query and loop through them to see what they represent.

wikidata_ids = ["Q36322","Q47461344","P569","P570","P31","P50","P577","P27","P21"]

for wikidata_id in wikidata_ids:

    linked_data_uri = f"https://www.wikidata.org/wiki/Special:EntityData/{wikidata_id}.json"

    response = requests.get(linked_data_uri, headers=wiki_auth_headers)
    data = response.json()

    entities = data.get('entities',{}).get(wikidata_id)
    labels = entities['labels']
    label = labels.get('en').get('value')

    descriptions = entities['descriptions']
    description = descriptions['en']["value"]

    print(f"{wikidata_id}: {label} - {description}")

Q36322: Jane Austen - English novelist (1775–1817)
Q47461344: written work - any work expressed in writing, such as inscriptions, manuscripts, documents or maps
P569: date of birth - date on which the subject was born
P570: date of death - date on which the subject died
P31: instance of - type to which this subject corresponds/belongs. Different from P279 (subclass of); for example: K2 is an instance of mountain; volcano is a subclass of mountain
P50: author - main creator(s) of a written work (use on works, not humans); use P2093 (author name string) when Wikidata item is unknown or does not exist
P577: publication date - date or point in time when a work or product was first published or released
P27: country of citizenship - the object is a country that recognizes the subject as its citizen
P21: sex or gender - sex or gender identity of human or animal. For human: male, female, non-binary, intersex, transgender female, transgender male, agender, etc. For animal: male organism, female organism. Groups of same gender use subclass of (P279)

Wikidata query service (SPARQL)

The Wikidata query service allows users to use SPARQL to retrieve information from Wikidata, meaning that more complex questions can be asked of the data. Similarly to the Linked data interface, it requires the Wikidata IDs to be known.

A simple query

To begin with, here is a simple SPARQL query taken from Alex's original that retrieves the birth and death dates for Jane Austen:

wikidata_query = """
    SELECT ?ja_birth ?ja_death WHERE {
      wd:Q36322 wdt:P569 ?ja_birth .
      wd:Q36322 wdt:P570 ?ja_death .
    }
  """

query_service_url = "https://query.wikidata.org/sparql"

params = {
    'format': 'json',
    'language': 'en',
    'query': wikidata_query
    }


response = requests.get(query_service_url, params=params, headers=wiki_auth_headers)
data = response.json()
print(data["results"]["bindings"])

[{'ja_birth': {'datatype': 'http://www.w3.org/2001/XMLSchema#dateTime', 'type': 'literal', 'value': '1775-12-16T00:00:00Z'}, 'ja_death': {'datatype': 'http://www.w3.org/2001/XMLSchema#dateTime', 'type': 'literal', 'value': '1817-07-18T00:00:00Z'}}]

The Service helper function

The SERVICE wikibase:label { bd:serviceParam wikibase:language "en" } line in Alex's query is a helper function that retrieves human-readable labels for the results. You can specify which language to retrieve the label for, and also specify other fields to retrieve (e.g. the description) too.

wikidata_query = """
    SELECT ?ja_birthLabel ?ja_deathLabel  WHERE {
      wd:Q36322 wdt:P569 ?ja_birth .
      wd:Q36322 wdt:P570 ?ja_death .
      SERVICE wikibase:label { bd:serviceParam wikibase:language "en" . }
    }
  """

query_service_url = "https://query.wikidata.org/sparql"

params = {
    'format': 'json',
    'language': 'en',
    'query': wikidata_query
    }


response = requests.get(query_service_url, params=params, headers=wiki_auth_headers)
data = response.json()
print(data["results"]["bindings"])

[{'ja_birthLabel': {'type': 'literal', 'value': '1775-12-16T00:00:00Z'}, 'ja_deathLabel': {'type': 'literal', 'value': '1817-07-18T00:00:00Z'}}]

The FILTER keyword

Allows the query to be filtered. In our example, the filter is applied to the publication date of any books published between the birth and death dates of Jane Austen.

wikidata_query = """
    SELECT ?author ?authorLabel  WHERE {
        wd:Q36322 wdt:P569 ?ja_birth .
        wd:Q36322 wdt:P570 ?ja_death .
        ?publication wdt:P31 wd:Q47461344 .
        ?publication wdt:P50 ?author .
        ?publication wdt:P577 ?pubdate .
        FILTER(?pubdate > ?ja_birth) .
        FILTER(?pubdate < ?ja_death) .
        SERVICE wikibase:label { bd:serviceParam wikibase:language "en" . }
    }
  """

query_service_url = "https://query.wikidata.org/sparql"

params = {
    'format': 'json',
    'language': 'en',
    'query': wikidata_query
    }


response = requests.get(query_service_url, params=params, headers=wiki_auth_headers)
data = response.json()

for binding in data["results"]["bindings"][:10]:
    print(binding)

{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q6527'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Jean-Jacques Rousseau'}}
{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q9235'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Georg Wilhelm Friedrich Hegel'}}
{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q1020553'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Miklós Küzmics'}}
{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q5879'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Johann Wolfgang von Goethe'}}
{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q9312'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Immanuel Kant'}}
{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q661076'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'John Hawkins'}}
{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q101638'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Mary Wollstonecraft'}}
{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q213681'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Adolph Freiherr Knigge'}}
{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q6722'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Carl Friedrich Gauss'}}
{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q28494'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Heinrich von Kleist'}}

Completing the query

For completeness, here is the full query that Alex originally created. Any country or gender properties associated with the filtered list of authors are also retrieved.

wikidata_query = """
    SELECT DISTINCT ?author ?authorLabel ?genderLabel ?countryLabel
    WHERE {
      wd:Q36322 wdt:P569 ?ja_birth .
      wd:Q36322 wdt:P570 ?ja_death .
      ?publication wdt:P31 wd:Q47461344 .
      ?publication wdt:P50 ?author .
      ?publication wdt:P577 ?pubdate .
      FILTER(?pubdate > ?ja_birth) .
      FILTER(?pubdate < ?ja_death) .
      ?author wdt:P27 ?country .
      ?author wdt:P21 ?gender
      SERVICE wikibase:label { bd:serviceParam wikibase:language "en" }
  }
"""

query_service_url = "https://query.wikidata.org/sparql"

params = {
    'format': 'json',
    'language': 'en',
    'query': wikidata_query
    }

response = requests.get(query_service_url, params=params, headers=wiki_auth_headers)
data = response.json()

for binding in data["results"]["bindings"][:10]:
    print(binding)

{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q6722'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Carl Friedrich Gauss'}, 'genderLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'male'}, 'countryLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Confederation of the Rhine'}}
{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q6722'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Carl Friedrich Gauss'}, 'genderLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'male'}, 'countryLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Kingdom of Hanover'}}
{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q6527'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Jean-Jacques Rousseau'}, 'genderLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'male'}, 'countryLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Republic of Geneva'}}
{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q661076'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'John Hawkins'}, 'genderLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'male'}, 'countryLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Kingdom of Great Britain'}}
{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q213681'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Adolph Freiherr Knigge'}, 'genderLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'male'}, 'countryLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Germany'}}
{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q101638'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Mary Wollstonecraft'}, 'genderLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'female'}, 'countryLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Kingdom of Great Britain'}}
{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q213681'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Adolph Freiherr Knigge'}, 'genderLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'male'}, 'countryLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Electorate of Hanover'}}
{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q9235'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Georg Wilhelm Friedrich Hegel'}, 'genderLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'male'}, 'countryLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Kingdom of Württemberg'}}
{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q9312'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Immanuel Kant'}, 'genderLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'male'}, 'countryLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Kingdom of Prussia'}}
{'author': {'type': 'uri', 'value': 'http://www.wikidata.org/entity/Q5879'}, 'authorLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Johann Wolfgang von Goethe'}, 'genderLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'male'}, 'countryLabel': {'xml:lang': 'en', 'type': 'literal', 'value': 'Saxe-Weimar-Eisenach'}}

A bonus picture round

One of the many benefits of working with Wikidata is that data from other Wikimedia platforms can be retrieved and incorporated into your queries. Wiki Commons https://commons.wikimedia.org/wiki/Main_Page is Wikimedia's platform for sharing open-access media files. Linkages can be made between Wikidata and Wiki Commons, meaning that it's often possible to retrieve images (Wikidata ID :P18) as part of a SPARQL query.

Here's a simple query that retrieves an image for Jane Austen:

from PIL import Image

wikidata_pic_query = """
  SELECT ?author ?pic
    WHERE 
    {
      wd:Q36322 wdt:P18 ?pic .
      wd:Q36322 wdt:P735 ?author .

    }
"""

query_service_url = "https://query.wikidata.org/sparql"

params = {
    'format': 'json',
    'language': 'en',
    'query': wikidata_pic_query
    }

response = requests.get(query_service_url, params=params, headers=wiki_auth_headers)

data = response.json()
bindings = data["results"]["bindings"]
image_url = bindings[0]['pic']['value']

# request image as bytes data
img = Image.open(requests.get(image_url, stream=True, headers=wiki_auth_headers).raw)

# resize image to 25% of original
width, height = img.size
width = int(width * .25)
height = int(height * .25)

img = img.resize((width, height))
print(image_url)
display(img)

http://commons.wikimedia.org/wiki/Special:FilePath/CassandraAusten-JaneAusten%28c.1810%29%20hires.jpg

	Type	Author(s)	Tags
Linking ACMI's collection data – Paul Duchesne's WikiFest 2025 presentation Connecting collections with Wikimedia->	documentation	Paul Duchesne->	Linked open data Open access Wikimedia
Combining Wikidata and GLAM data – Tim Sherratt's WikiFest 2025 presentation Connecting collections with Wikimedia->	documentation	Tim Sherratt->	Open access Linked open data
How linked open data helped build an exhibition – Julia Rodwell's WikiFest 2025 presentation Connecting collections with Wikimedia->	documentation	Julia Rodwell->	Exhibitions Wikimedia Digital culture Linked open data
What is Wikidata? Alex Lum's WikiFest 2025 presentation Connecting collections with Wikimedia->	documentation	Alex Lum->	Linked open data Open access Wikimedia

How to query Wikidata

How to query Wikidata

What is SPARQL?

Triples format

Introducing SPARQL (and Wikidata) syntax

Querying Wikidata

Authorising requests to Wikimedia

Search

Linked data interface

Wikidata query service (SPARQL)

A simple query

The Service helper function

The FILTER keyword

Completing the query

A bonus picture round

Further reading + useful links

Related Resources