SPARQL/Triples

Introduction The statement "The sky has the color blue", consists of a subject ("the sky"), a predicate ("has the color"), and an object ("blue"). SPO or "subject, predicate, object" is known as a (Semantic) triple, or commonly referred to in Wikidata as a statement about data. SPO is also used as a form of basic syntax layout for querying RDF data structures, or any graph database or triplestore, such as the Wikidata Query Service (WDQS). See also w:en:Semantic triple In Wikidata Query Service (WDQS) triples are used to describe the Query pattern in the WHERE clause of the SELECT statement # ?child father Bach ?child wdt:P22 wd:Q1339. In this case the triple ?child wdt:p22 wd:Q1339 specifies that the variable ?child must have the parent/father Bach. Any of the triple parts Subject, Predicate and Object may be variables. This makes this selection very versatile. Triples with the same subject 500px|right|Example of SPARQL Triples Additional variables can be added by adding additional triples. In the simplest case these triples use the same subject. SELECT ?child ?childLabel ?genderLabel ?birth_date ?date_of_death WHERE { ?child wdt:P22 wd:Q1339.# ?child has father Bach ?child wdt:P21 ?gender. ?child wdt:P569 ?birth_date. ?child wdt:P570 ?date_of_death. SERVICE wikibase:label { bd:serviceParam wikibase:language "en". } } The first triple selects all the children of Bach. The additional triples links all these triples with a value for gender, birth date and date of death. The variable ?child links all of them together. If you look closely at the result you might have noticed that Johann Christoph Friedrich Bach has 2 lines in the list because there are 2 different birth dates, 21 and 23 of June 1732. In his case ?child wdt:P569 ?birth_date. resulted into 2 values. See for further details at removing duplicates and modifiers. OPTIONAL triples If not all subjects have a value for a certain triple the subject is excluded. To have it included the OPTIONAL keyword comes in handy. SELECT DISTINCT ?child ?childLabel ?genderLabel ?birth_date ?date_of_death WHERE { ?child wdt:P22 wd:Q76.# ?child has father Obama OPTIONAL{ ?child wdt:P21 ?gender. } OPTIONAL{ ?child wdt:P569 ?birth_date. } OPTIONAL{ ?child wdt:P570 ?date_of_death. } SERVICE wikibase:label { bd:serviceParam wikibase:language "en". } } Both children are shown, even if one of the variables (in this case the date of death) is not filled in. See the chapter OPTIONAL for a full description. Complex triples Triples are not limited to one subject. In fact triples can be linked in any thinkable way. You would for instance be able to list the coordinates of the birth places of the children of Bach SELECT ?child ?childLabel ?placeofbirthLabel ?coordinates WHERE { ?child wdt:P22 wd:Q1339.# ?child has father Bach ?child wdt:P19 ?placeofbirth. ?placeofbirth wdt:P625 ?coordinates. SERVICE wikibase:label { bd:serviceParam wikibase:language "en". } } ORDER BY ?placeofbirthLabel You could even see these birthplaces (Köthen, Leipzig and Weimar) on a map by using #defaultView:Map defaultView:Map SELECT ?placeofbirthLabel ?coordinates (GROUP_CONCAT(DISTINCT ?childLabel; SEPARATOR=", ") AS ?children) WHERE { ?child wdt:P22 wd:Q1339.# ?child has father Bach ?child wdt:P19 ?placeofbirth. ?placeofbirth wdt:P625 ?coordinates. SERVICE wikibase:label { bd:serviceParam wikibase:language "en". ?child rdfs:label ?childLabel. ?placeofbirth rdfs:label ?placeofbirthLabel. } } GROUP BY ?placeofbirthLabel ?coordinates ?children If you click on a red dot you will get additional data as specified above with the variables ?placeofbirthLabel and ?children. We had to use GROUP BY, GROUP_CONCAT, DISTINCT and all labels should be defined explicitly in the SERVICE. You can toggle between the Map display and standard table display by the Display drop down list, at the right side of the Run button. See more about views at Map views or all views Triples by number of variables Triples with one variable An example of a triple with one variable for Subject would be SELECT ?child ?childLabel WHERE { ?child wdt:P22 wd:Q1339. # ?child has father Bach SERVICE wikibase:label { bd:serviceParam wikibase:language "en". } } This will list all Subjects (as variable ?child) with Predicate 22 and Object 1339. An example of a triple with one variable for Predicate would be SELECT ?predicate ?pLabel WHERE { wd:Q57225 ?predicate wd:Q1339. # Johann Christoph Friedrich Bach ?predicate Johann Sebastian Bach BIND( IRI(REPLACE( STR(?predicate),"prop/direct/","entity/" )) AS ?p). # or ?p wikibase:directClaim ?predicate. SERVICE wikibase:label { bd:serviceParam wikibase:language "en". } } This will list all Predicates (as variable ?predicate) with Object 57225 and Subject 1339. It shows that he is not only his 22 but also 1066 him An example of a triple with one variable for Object would be SELECT ?workloc ?worklocLabel WHERE { wd:Q1339 wdt:P937 ?workloc. # Bach work location ?workloc SERVICE wikibase:label { bd:serviceParam wikibase:language "en". } } This will list all Objects (as variable ?workloc) with Subject 1339 and Predicate 937. Triples with two variables An example of a triple with 2 variables and only a fixed value for Subject would list all raw information available in Wikidata about Bach SELECT ?predicate ?object WHERE { wd:Q1339 ?predicate ?object. # Bach } See further at next section with 3 variables for further usage An example of a triple with 2 variables and only a fixed value for Predicate would list all subjects (probably airports) with an IATA airport code SELECT ?subject ?subjectLabel ?object WHERE { ?subject wdt:P238 ?object. # IATA airport code SERVICE wikibase:label { bd:serviceParam wikibase:language "en". } } ORDER BY ?object An usage could be to check for duplicate IATA codes: SELECT ?object (COUNT(?subject) AS ?count) (MIN(?subject) AS ?subject1) (MAX(?subject) AS ?subject2) (GROUP_CONCAT(DISTINCT ?subjectLabel; SEPARATOR=", ") AS ?subjectLabels) WHERE { ?subject wdt:P238 ?object. # IATA airport code SERVICE wikibase:label { bd:serviceParam wikibase:language "en". ?subject rdfs:label ?subjectLabel. } } GROUP BY ?object HAVING(COUNT(?subject) > 1) ORDER BY ?object An example of a triple with 2 variables and only a fixed value for Object would list all subjects related to Bach SELECT ?subject ?subjectLabel ?subjectDescription ?predicate ?pLabel WHERE { ?subject ?predicate wd:Q1339. # Bach BIND( IRI(REPLACE( STR(?predicate),"prop/direct/","entity/" )) AS ?p). # or ?p wikibase:directClaim ?predicate. SERVICE wikibase:label { bd:serviceParam wikibase:language "en". } } ORDER BY ?subject An other possibility of a triple with fixed value for Object would list all subjects with value "ABC", and will show for instance airport Albacete Airport SELECT ?subject ?subjectLabel ?subjectDescription ?predicate ?pLabel WHERE { ?subject ?predicate "ABC". BIND( IRI(REPLACE( STR(?predicate),"prop/direct/","entity/" )) AS ?p). # or ?p wikibase:directClaim ?predicate. SERVICE wikibase:label { bd:serviceParam wikibase:language "en". } } ORDER BY ?subject Triples with three variables When you would use triples with all 3 as variables (one for Subject, one for Predicate and one for Object) you basically will list out the whole database. This can be done for small databases, and can be used as well to get a rough idea of the available data, on all available properties. All raw information available in Wikidata about the children of Bach: SELECT ?subject ?predicate ?object WHERE { ?subject ?predicate ?object. ?subject wdt:P22 wd:Q1339. # subject has father Bach } ORDER BY ?subject ?predicate ?object LIMIT 10000 The same query but grouped by predicate: SELECT DISTINCT ?subject ?subjectLabel ?predicate (GROUP_CONCAT(DISTINCT ?object; SEPARATOR=", ") AS ?objects) WHERE { ?subject ?predicate ?object. ?subject wdt:P22 wd:Q1339. # subject has father Bach SERVICE wikibase:label { bd:serviceParam wikibase:language "en". } } GROUP BY ?subject ?subjectLabel ?predicate ORDER BY ?subject ?subjectLabel ?predicate LIMIT 10000 From the query below you can discover triples about the date the Wikidata page was last updated, the total number of statements, the number of sitelinks etc. These are schema:dateModified, wikibase:statements and wikibase:sitelinks respectively. SELECT ?subject ?subjectLabel ?datemodified ?statements ?sitelinks WHERE { ?subject wdt:P22 wd:Q1339. # subject has father Bach ?subject schema:dateModified ?datemodified. ?subject wikibase:statements ?statements. ?subject wikibase:sitelinks ?sitelinks. SERVICE wikibase:label { bd:serviceParam wikibase:language "en". } }