Wikidata SPARQL endpoint in Mathematica

Wikidata SPARQL endpoint in Mathematica#

by Vishank Patel

See the Wikidata Welcome Page, Wikidata Query Service/User Manual and Mathematica WikidataData documentation.

These recipe examples were tested on February 10, 2022.

1. Basic Wikidata request#

Mathematica has a built-in function “WikidataData” to request information from WikiData’s SPARQL endpoint. Arguments can be in the form of Mathematica entities or external identifiers, which are unique for every listed object.

Mathematica Entities#

Mathematica entities can be queried using Mathematica’s free form input (“Ctrl” + “ =” , then typing the name in normally). As we are in a jupyter notebook, we will need to use the Entity function:

Entity["Planet", "Earth"]["Radius"]
Output

External identifiers#

Wikidata#

External identifiers from Wikidata can be selected using the WikidataSearch command.

WikidataSearch["Earth"]
Output

As you can see, we get a list of outputs from Wikidata. To extract the first element, we use the shorthand notation for the part command:

WikidataSearch["Earth"][[1]]
Output

To extract information about the object, we query WikidataData using the identifier

WikidataData[WikidataSearch["Earth"][[1]], "Radius"]
Output

WikidataData’s syntax is generally the following:

WikidataData[item, {property1, property2, …}]

To check all the available properties for an entity, we can run the following command:

WikidataData[WikidataSearch["Nick Saban"][[1]], "Properties"]
Output

If the properties requested are more specialized, they can be inputted by searching for the property object on Wikidata.

WikidataData[
 WikidataSearch["Nick Saban"][[1]],
 WikidataSearch["Property" -> "Award received"][[1]]]
Output

Other Identifiers#

Wikidata can also accept input from other identifier systems such as ISBN, DOI, PubChemCompoundID and many more:

WikidataData[
 ExternalIdentifier["PubChemCompoundID", "123591"],
 WikidataSearch["Property" -> "Chemical structure"][[1]]]
Output

2. Example Use-cases#

Entity classes behave like a collection of entities. Defining classes allows us to add constrains to our search and get specific useful data.

Retrieving melting points of elements#

WikidataData[
    {Entity["Element","Aluminium"],Entity["Element","Lead"],Entity["Element","Tungsten"]},
    WikidataSearch["Property"->"melting point"][[1]],"Association"]
Output

To be more specific, we can obtain the melting points of elements with density from 7 to 8 grams/cm^3

WikidataData[
 EntityList[
  EntityClass[
   "Element", 
   {"Density" -> Quantity[Interval[{7, 8}], "Grams"/"Centimeters"^3]}]],
 WikidataSearch["Property" -> "melting point"][[1]], "Association"]
Output

NOTE : In Mathematica, the EntityClass above can be easily constructed using Mathematica’s free-form input. For instance, typing “earliest discovered elements” in the orange box(“Ctrl”+”=”) gives the following output:

WikidataData[
 EntityList[
  EntityClass[
   "Element", {EntityProperty["Element", "DiscoveryDate"] -> TakeLargest[5]}]],
 WikidataSearch["Property" -> "discoverer"][[1]], "Dataset"]
Output

Population distribution of Alabama counties#

populationData =
  WikidataData[
   EntityList[
    Entity["AdministrativeDivision", {"Alabama", "UnitedStates"}]["Subdivisions"]],
   WikidataSearch["Property" -> "population"][[1]], 
   "Association"];

ReverseSortBy[populationData, N] // Shallow
Output

Publication dates of books written by Newton#

WikidataData[
 EntityClass[
  WikidataSearch["Written work"][[1]], 
  {WikidataSearch["Property" -> "Author"][[1]] -> Entity["Person", "IsaacNewton"]}],
 WikidataSearch["Property" -> "publication date"][[1]], 
 "Dataset"]
Output

SI unit symbols and their measured physical quantities#

WikidataData[
 EntityClass[WikidataSearch["SI base Unit"][[1]], {}],
 {WikidataSearch["Property" -> "unit symbol"][[1]],
  WikidataSearch["Property" -> "Measured physical quantity"][[1]]},
  "Dataset"]
Output

3. SPARQL Query Syntax in Mathematica#

Importing the necessary packages:

Needs["GraphStore`"]

It is better to write the queries in the WikiData Query Service (WDQS) at https://query.wikidata.org/ given the autocompletion functionality.

WikiData SPARQL tutorial: https://www.wikidata.org/wiki/Wikidata:SPARQL_tutorial (It is amazing!)

Once the query is written, we can run it in Mathematica as follows:

US Renewable Energy Farms#

endpoint = "https://query.wikidata.org/sparql";
WindQuery = "SELECT ?powerplant ?pos ?powerplantLabel
WHERE
{
  ?powerplant wdt:P17 wd:Q30;
              wdt:P31 wd:Q194356;
              wdt:P625 ?pos;
  SERVICE wikibase:label { bd:serviceParam wikibase:language \"[AUTO_LANGUAGE],en \". }
}";

WindFarmData = SPARQLExecute[endpoint, WindQuery];
WindFarmData //Shallow
Output
GeoListPlot["pos" /. WindFarmData]
Output
SolarQuery = "SELECT ?powerplant ?pos ?powerplantLabel
WHERE
{
  ?powerplant wdt:P17 wd:Q30;
              wdt:P31 wd:Q1003207;
              wdt:P625 ?pos;
  SERVICE wikibase:label { bd:serviceParam wikibase:language \"[AUTO_LANGUAGE],en\". }
}";

SolarFarmData = SPARQLExecute[endpoint, SolarQuery];
SolarFarmData // Shallow
Output
GeoListPlot["pos" /. SolarFarmData]
Output
Contents