- All important points from the training course, collected for quick reference. See also [[Tutorial]]s. - ### SPARQL Interactions id:: 626e4952-02cc-4c88-8565-139a93e6a98a - ![SPARQLinteractions.svg](../assets/SPARQLinteractions_1643113721643_0.svg){:height 304, :width 746} - See [[SPARQL Editors]] - ### Two triple patterns, same subject id:: 626e4952-6ffd-494a-bbbf-122685d3d4f7 - ```sparql SELECT ?country ?capital ?population WHERE { ?country dbo:capital ?capital . ?country dbp:populationTotal ?population . } ``` #Query #DBpedia - ![TwoPatternsSameSubject.svg](../assets/TwoPatternsSameSubject_1643114435930_0.svg){:height 279, :width 522} - ### Making another step in the graph id:: 626e4952-98ae-4115-93de-d9fd4156978c - ```sparql SELECT ?country ?population ?capital WHERE { ?country a dbo:Country . ?country dbp:populationTotal ?population . ?country dbo:capital ?capital_URI . ?capital_URI rdfs:label ?capital . } ``` #Query #DBpedia - ![AnotherStepInTheGraph.svg](../assets/AnotherStepInTheGraph_1643114924911_0.svg){:height 373, :width 646} - collapsed:: true ```sparql PREFIX skos: PREFIX skosxl: SELECT * WHERE { ?s skosxl:prefLabel ?Label_URI . ?Label_URI skosxl:literalForm ?label . FILTER (LANG(?label) = "es") } ``` - #VocBench #Query - ### Three patterns and two filters collapsed:: true id:: 626e4952-606b-45fe-8878-dfe4871a59fc - ```sparql SELECT ?country ?population ?capital WHERE { ?country a dbo:Country; dbo:populationTotal ?population; dbo:capital ?capital_URI . ?capital_URI rdfs:label ?capital . FILTER ( LANG(?capital) = "en" ) FILTER ( xsd:int(?population) > 10000000 ) } ORDER BY DESC(?population) ``` #Query #DBpedia #FILTER #LANG - ### Counting and Binding collapsed:: true id:: 626e4952-dd36-4931-abc1-98e40bb1da76 - ```sparql SELECT (COUNT (?work_of_Agatha_Christie) AS ?Number_of_AC_works) {?work_of_Agatha_Christie dbo:author dbr:Agatha_Christie } ``` #Query #COUNT - ### Functions on dates and times collapsed:: true id:: 626e4952-30dc-4bce-804c-fee5318b8244 - [[NOW]]() - [[YEAR]]() - [[MONTH]]() - [[DAY]]() - [[HOURS]]() - [[MINUTES]]() - [[SECONDS]]() - [[Example]] collapsed:: true - ```sparql SELECT (MONTH(NOW()) AS ?currentMonth) { } ``` #Query - ### Filtering resources with the same namespace id:: 626e4952-f335-485a-b3ae-474fb7777046 - Find all [[SKOS]] properties in use - ```sparql SELECT DISTINCT ?property WHERE { ?s ?property ?o . FILTER (STRSTARTS(STR(?property), "http://www.w3.org/2004/02/skos/core#" )) } ``` #Query #STRSTARTS - Find all classes of things that have author DBpedia - ```sparql SELECT DISTINCT ?class WHERE { ?work dbo:author ?author . ?work a ?class . FILTER (STRSTARTS(STR(?class),"http://dbpedia.org/ontology/" )) } ``` #Query #STRSTARTS #DBpedia - ### Negation collapsed:: true id:: 626e4952-7079-4ddb-a19d-72eb4dc7cca2 - `NOT EXISTS` - Testing for absence of a pattern (recommended in most cases) - `MINUS` - Removing possible solutions - `!BOUND` - Not bound (produces the same results as `NOT EXISTS` but works also for SPARQL 1.0) - #+BEGIN_IMPORTANT `NOT EXISTS` and `MINUS` have different logic and may produce different results on the same dataset. #+END_IMPORTANT - ```sparql SELECT DISTINCT ?leader WHERE { ?MS a dbo:Country; dct:subject dbc:Member_states_of_the_European_Union; dbp:leaderName ?leader . FILTER NOT EXISTS {?leader dbo:spouse ?spouse .} } ORDER BY DESC(?leader) ``` #Query #FILTER #[[NOT EXISTS]] - ### Optional results collapsed:: true id:: 626e4952-6483-438f-9a57-c210fdd6988a - ```sparql SELECT DISTINCT ?leader ?spouse WHERE { ?MS a dbo:Country; dct:subject dbc:Member_states_of_the_European_Union; dbp:leaderName ?leader . OPTIONAL {?leader dbo:spouse ?spouse .} } ``` #Query #OPTIONAL - ### Combining results collapsed:: true id:: 626e4952-211a-4446-b227-873a647fe005 - ```sparql SELECT DISTINCT ?Influencer ?Influenced WHERE { {dbr:Ludwig_Wittgenstein dbo:influencedBy ?Influencer .} UNION {dbr:Ludwig_Wittgenstein dbo:influenced ?Influenced . } } ``` #Query #UNION - #+BEGIN_NOTE This query will not give complete results, unless inferred triples on inverse properties are included. To guarantee complete results relying only on SPARQL, the inverse property should be included. #+END_NOTE - ### Search for string patterns id:: 61fd3a3d-81b5-448d-a76f-d1b0008b7476 collapsed:: true - ```sparql SELECT DISTINCT ?MusicalWork ?Artist WHERE { ?MusicalWorkURI a dbo:MusicalWork ; rdfs:label ?MusicalWork ; dbo:artist ?ArtistURI . ?ArtistURI rdfs:label ?Artist . FILTER (lang(?MusicalWork) = "en") FILTER (lang(?Artist) = "en") FILTER REGEX (?MusicalWork, "broken", "i") } ORDER BY ?MusicalWork ``` #Query #REGEX #FILTER #LANG - ### Assembling strings collapsed:: true id:: 626e4952-d6a2-41d0-9766-10286789ae0c - ```sparql PREFIX cdm: SELECT ?reg (IRI(CONCAT(STR(?eli),"/eng/pdf")) AS ?pdf_URL) { ?reg cdm:resource_legal_eli ?eli . } LIMIT 1000 ``` #Query #IRI #CONCAT #Cellar - ### Property paths collapsed:: true id:: 626e4952-c45c-4911-bda6-6c93293dbe9e - [[Syntax]] collapsed:: true - ```sparql #Inverse Path: ^:property #Inverse path (object to subject) #Sequence Path: :property1/:property2 #A sequence path of :property1 followed by :property2 #Alternative Path: :property1|:property2 #:property1 or property2 (all possibilities are tried) #Zero Or More: :property* #A path that connects the subject and object of the path by zero or more matches of :property #One Or More: :property+ #A path that connects the subject and object of the path by one or more matches of :property #Zero Or One Path: :property? #A path that connects the subject and object of the path by zero or one matches of :property ``` - [Property paths from the SPARQL 1.1 spec](https://www.w3.org/TR/sparql11-query/#propertypaths) - [[Example]] collapsed:: true - ```sparql SELECT (COUNT (DISTINCT ?child) AS ?Number_of_children) { ?MS a dbo:Country; dct:subject dbc:Member_states_of_the_European_Union ; dbp:leaderName ?leader . ?leader dbo:child|^dbo:parent ?child. #Works the same as # {?leader dbo:child ?child.} # UNION # {?child dbo:parent ?leader} } ``` #Query #| #^ #UNION - ![PropertyPath_ChildParent.svg](../assets/PropertyPath_ChildParent_1645774371298_0.svg) - ```sparql SELECT (COUNT (DISTINCT ?Influencer) AS ?Num_Influencer) (COUNT (DISTINCT ?Influenced) AS ?Num_Influenced) WHERE { {dbr:Ludwig_Wittgenstein dbo:influencedBy|^dbo:influenced ?Influencer .} UNION {dbr:Ludwig_Wittgenstein dbo:influenced|^dbo:influencedBy ?Influenced .} } ``` #Query #DBpedia #| #^ #COUNT #UNION - - ### Language tag collapsed:: true id:: 626e4952-9418-419f-918a-47f28f524f66 - The current best practice for language tags is defined in [RFC5646](https://www.rfc-editor.org/rfc/rfc5646.txt). - [[LANGMATCHES]] collapsed:: true - `FILTER (LANG(?label) = "fr")` will only return results where the language tag of `?label` is `@fr`, while `FILTER LANGMATCHES( LANG(?label), "fr" )` will also include those with regions such as `@fr-BE`. - ### Wikibase data model collapsed:: true id:: 626e4952-c934-4fa9-a7e1-eb3803156299 - ![image.png](../assets/image_1646977222963_0.png) - ### Wikidata label service collapsed:: true id:: 626e4952-5283-4052-a77e-c12132783160 - Adding "Label" to a variable will get the label of the respective item if the Wikidata Label service is called in the query. For example, for ?item, ?itemLabel will get the labels in the selected language. - ```sparql SELECT ?item ?itemLabel WHERE { ?item wdt:P31 wd:Q146 . SERVICE wikibase:label { bd:serviceParam wikibase:language “en" } } ``` #Query #Wikidata - #+BEGIN_TIP To include a call to the Label Service in your query, just press Ctrl-Space and start typing "label" until you see the query fragment. Then change the language if needed. Check the [Label service manual](https://www.mediawiki.org/wiki/Wikidata_Query_Service/User_Manual#Label_service) for other capabilities. #+END_TIP - ### Wikidata Language and visualization service collapsed:: true id:: 626e4952-d248-4ef9-8f71-5b0cb5a0f824 - Influence graph in the age of Enlightenment #Wikidata #Query #FILTER #&& - ### Properties for statements (Wikidata) id:: 600748be-0c7a-4b4c-807d-a5c41f8ee80e collapsed:: true - Item - Statement - Value - Qualifier Value collapsed:: true - ![WikidataStatementPropertiesGeneric.svg](../assets/WikidataStatementPropertiesGeneric_1644054801681_0.svg) - [[Example]]: US Presidents collapsed:: true - - ![WikidataStatementPropertiesUSpresident.svg](../assets/WikidataStatementPropertiesUSpresident_1644055097874_0.svg){:height 502, :width 680} - With `?statement` variable #Query #Wikidata #Qualifiers - With blank nodes #Query #Wikidata #Qualifiers - See also ((61fe46bb-6fc5-4d5e-b917-d0da5b02a96a)) - ### Providing list of input VALUES collapsed:: true id:: 626e4952-8b32-4fc4-b3a2-6e24c7824ada - ```sparql # Example with a block of values, where the third block in the list has no value for ?v2 VALUES (?v1 ?v2) {(:c11 :c21) (:c21 :c22) (:c21 UNDEF)} # When value for only one variable are provided, () can be omitted VALUES ?v1 {:c1 :c2 :c3 :cN} ``` #VALUES - Get the list of notable books on linked data, semantic technologies, and semantic knowledge graphs. - - ### Grouping results with GROUP_CONCAT collapsed:: true id:: 626e4952-ad6c-4838-bab3-c76ce434120a - ```sparql (GROUP_CONCAT (?V; SEPARATOR = ", ") AS ?groupV) ``` #GROUP_CONCAT - ### Federated queries collapsed:: true id:: 626e4952-1bba-4499-b273-3034b07779b4 - ![FederatedQueries.svg](../assets/FederatedQueries_1650444816370_0.svg){:height 500, :width 643} - ### Default and named graphs id:: 626e4952-3b13-47d8-a7e4-ad95ca6b68e2 collapsed:: true - ![DefaultAndNamedGraph.svg](../assets/DefaultAndNamedGraph_1650444948710_0.svg){:height 454, :width 498} - Default graph `FROM <>` Used when not explicitly specified - Named graph `FROM NAMED <>` Can be directly specified or through a variable - Available named graphs at an endpoint - ```sparql SELECT DISTINCT ?g WHERE {    GRAPH ?g {?s ?p ?o .} } ``` #Query #GRAPH - ### Other query forms id:: 626e4952-fb6b-4afc-8dac-d0bec45a3a8e collapsed:: true - ASK - If there is at least one result - `ASK WHERE {...}` - DESCRIBE - Describes a particular resource - All directly linked nodes - Not strictly specified - [[CONSTRUCT]] - Create new graph from a template - Use SPARQL as template to - Copy - Create - Convert - From RDF to RDF - Applying different ontologies - Generating URIs - From CSV to RDF (e.g. using [Tarql](http://tarql.github.io)) - Result - Triples (virtual graph) - [[Example]] - SKOS from SKOS-XL - ```sparql CONSTRUCT {     ?concept skos:prefLabel ?prefLabel_ES .} WHERE { ?concept skosxl:prefLabel ?prefLabel . ?prefLabel skosxl:literalForm ?prefLabel_ES . FILTER (LANG(?prefLabel_ES) = "es") } ``` #Query #CONSTRUCT #FILTER #LANG