Lecture 10: Storing and querying RDF data

UNIVERSITY OF JYVÄSKYLÄ 

Lecture 10: Storing and querying RDF data 

University of Jyväskylä 

ITKS544 Semantic Web and Ontology Engineering 

Spring 2012 

Version 1.0 

Michal Nagy


Environmental variables 

� OS stores simple variables in its memory 

� Applications can ask OS for the value 

� Each variable has a name and a simple textual value 

� To see variables and their values run (Win): 

– See all: 

– See a concrete variable: 

� Scope: 

set 

– Global (valid for the whole OS) 

– Local (valid in the current command line window) 

� To change or set new local variable run: 

set VARNAME=some text here 

set VARNAME 

practical 

27/02/2012 ITKS544 - Lecture 10 2


� Plain literals 

– Represent strings 

– Optional language tag 

– Example: “John”@en 

� Typed literals 

More about literals 

– Represent non-string values (e.g. boolean, numbers, …) 

– Lexical form of the literal + URI of the datatype 

– Example: “true”^^ 

@prefix f: . 

@prefix xsd: . 

f:john f:age "25"^^xsd:int ; 

f:firstname "John"@en ; 

f:taxesPayed "true"^^xsd:boolean ; 



Storing RDF data 



Storing of RDF 

� Small datasets (few triples) 

– RDF file published on the web or stored locally 

� Large datasets (thousands to millions of triples) 

– Database-bases solution better 

– Usually in form of RDF storage 

� Legacy data 

– Keep in original form 

– Provide mapping to RDF 

– Expose as RDF to the outer world 



� Jena (open-source) 

Database-based solutions 

– Link: http://incubator.apache.org/jena/ 

� Joseki (open-source) 

– Link: http://www.joseki.org/ 

� Sesame by Aduna (open-source) 

– Link: http://www.openrdf.org/ 

� RDFLib (open source, BSD 2-Clause license) 

– Link: http://code.google.com/p/rdflib/ 

– Doc: http://readthedocs.org/docs/rdflib/en/latest/ 

� Mulgara (open source, Open Software License) 

– Link: http://www.mulgara.org/ 

� Open Anzo (open source, Ecplise Public License 1.0) 

– Link: http://www.openanzo.org/ 

� Virtuoso Universal Server by OpenLink Software 

– Link: http://virtuoso.openlinksw.com/ 



Sesame 

� Framework for processing RDF data 

� Home: http://www.openrdf.org/ 

� Features: 

– Parsing 

• Supports all major notations 

– Storing 

• In-memory, RDBS-backed, file-based 

– Inferencing 

• Rule-based, Ontology-based 

– Querying 

• SPARQL, SeRQL 

� Java-based API + tools 

� Elmo: http://semanticweb.org/wiki/Elmo 



Sesame architecture 

� Rio (RDF I/O) 

– Parsers and writers for various notations 

� Sail (Storage And Inference Layer) 

– Low level System API 

– Abstraction for storage and inferencing 

� Repository API 

– Higher level API 

– Developer-oriented methods for handling RDF data 

� HTTP Server 

– Accessing Sesame through HTTP 

Image source: http://www.openrdf.org/doc/sesame2/users/ch03.html 



Installing Sesame workbench 

� Simple web interface for storing and querying RDF data 

� Install steps (no admin rights needed): 

practical 

1. Download and unzip newest Sesame and Tomcat 

2. Copy all *.war files from Sesame’s war folder to Tomcat’s 

webapps folder 

3. Start Tomcat 

• From bin folder by running startup.sh (UNIX) or 

startup.bat (Win) 

• You may need to set JAVA_HOME variable (it should point to 

JDK or JRE main folder) 

4. Go to http://localhost:8080/openrdf-workbench 



Jena 

� Collection of tools and Java libraries to simplify the 

development of semantic web and linked-data 

applications 

� Home: http://incubator.apache.org/jena/index.html 

� Includes: 

– API for processing RDF data in various notations 

– Ontology API for OWL and RDFS 

– Rule-based inference engine 

– RDF store 

– SPARQL query processor (called ARQ) 

27/02/2012 ITKS544 - Lecture 10 10


� Installation: 

– Download and unzip Jena 

ARQ 

– Set JENAROOT environmental variable to folder where you 

unzipped Jena 

– In bat (Win) or bin (UNIX) you find arq executable 

� Usage: 

practical 

– Prepare a SPARQL query and save it into a file (here: 

query.sparql) 

– Prepare some data file (if needed) – e.g. data.rdf 

– Execute the query on top of the data by running: 

arq --query=query.sparql --data=data.rdf 

� In case of problems use arq --help 



� SPARQL server for Jena 

� Features: 

Joseki 

– RDF Data from files and databases 

– HTTP (GET and POST) implementation of the SPARQL 

protocol 

� Installation: 

practical 

1. Download and unzip Joseki 

2. Check if you can run java from command line 

• If not, put java to PATH environmental variable 

3. Start Joseki server from Joseki folder by running: 

java -cp "%CLASSPATH%;lib\*" joseki.rdfserver joseki-config.ttl 

4. Open http://localhost:2020/ 



� Python-based 

RDFlib 

� Link: http://code.google.com/p/rdflib/ 

� Documentation: http://readthedocs.org/docs/rdflib/ 

� Simple control over RDF graphs within Python code 

– Parsing and serializing 

� Storing of RDF data using relational DB 

� Querying using SPARQL 



RDFlib plugins 

Image source: http://readthedocs.org/docs/rdflib/en/latest/ 



� RDF database 

� Written entirely in Java 

� Link: http://www.mulgara.org/ 

Mulgara 

� Querying language – SPARQL and own TQL 

� TQL language: 

– Interpreted querying and command language 

– To manage Mulgara storage 

� REST interface for TQL and SPARQL 

� Starting the server: 

practical 

– Download and unzip the binary package 

– Inside the mulgara folder run: 

– TQL REST interface: http://localhost:8080/tql/ 

java -jar mulgara-2.1.13.jar 

– SPARQL REST interface: http://localhost:8080/sparql/ 



Querying RDF data 



@prefix f: 

. 

@prefix xsd: 

. 

f:john f:age "25"^^xsd:int . 

f:bill f:age "30"^^xsd:int . 

f:mary f:age "24"^^xsd:int . 

f:jane f:age "26"^^xsd:int . 

f:john f:loves f:mary . 

f:bill f:loves f:jane . 

f:john f:hasFriend f:bill 

Example data set 

f:bill 

f:john 

f:mary 

27/02/2012 ITKS544 - Lecture 10 17 

30 

25 

f:age 

f:age 

f:loves 

f:loves 

f:jane 

26 

24 

f:age 

f:age


Simple SPARQL queries (1) 

� Show me the property f:age of resource f:mary 

f:bill f:loves 

30 

f:john 

25 

f:age 

f:age 

Data 

f:loves 

f:jane 

26 

f:mary 

24 

f:age 

f:age 

Query 

SELECT ?a 

WHERE { 

?a } 

PREFIX f: 

SELECT ?a 

WHERE { f:mary f:age ?a } 

f:mary 


?a 

f:age 

Result 

a 

24



� Show me f:age of all resources 


30 

f:john 

25 

f:age 

f:age 

Data Query 

f:loves 

f:jane 

26 

f:mary 

24 

f:age 

f:age 

PREFIX f: 

SELECT ?r ?a 

WHERE { ?r f:age ?a } 

Result 

r a 

f:bill 30 

f:jane 26 

f:john 25 

f:mary 24 


?r 

?a 

f:age



� Show me all things that are loved. Also show me their age (f:age) 

?x 


30 

f:john 

25 

f:age 

f:age 

f:loves 

f:loves 

Data Query 

f:jane 

26 

f:mary 

24 

?t 

?a 

f:age 

f:age 

f:age 

PREFIX f: 

SELECT ?t ?a 

WHERE { 

?x f:loves ?t . 

?t f:age ?a 

} 

Result 

t a 

f:jane 26 

f:mary 24 



SPARQL: Optional condition 

� Show me the person and its age (f:age). If you have information about 

that person loving somebody, then show it as well. 

?p 

?a 


30 

f:john 

25 

f:age 

f:age 

f:age 

f:loves 

f:loves 

Data Query 

f:jane 

26 

f:mary 

24 

?q 

f:age 

f:age 

PREFIX f: 

SELECT ?p ?a ?q 

WHERE { 

?p f:age ?a . 

OPTIONAL {?p f:loves ?q} 

} 

Result 

p a q 

f:bill 30 f:jane 

f:john 25 f:mary 

f:mary 24 

f:jane 26 



SPARQL: Filtering 

� Show me people and their age for people older than 25. 


30 

f:john 

25 

f:age 

f:age 

Data Query 

f:loves 

f:jane 

26 

f:mary 

24 

f:age 

f:age 

PREFIX f: 

SELECT ?p ?a 

WHERE { 

?p f:age ?a . 

FILTER (?a > 25) 

} 

If ?a is not a number, then it will not work 

Result 

p a 

f:bill 30 

f:jane 26 



SPARQL: Logical OR (union) 

� Show me all people who have a friend together with all the people that 

are younger than 25 


30 

f:john 

25 

f:age 

f:age 

Data Query 

f:loves 

f:jane 

26 

f:mary 

24 

f:age 

f:age 

PREFIX f: 

SELECT ?p 

WHERE {?p f:age ?a . FILTER (?a < 25)} 

+ 

PREFIX f: 

SELECT ?p 

WHERE {?p f:hasFriend ?q} 

PREFIX f: 

SELECT ?p 

WHERE { 

{?p f:age ?a . FILTER (?a < 25)} 

UNION {?p f:hasFriend ?q} 

} 

Result 


p 

f:mary 

f:john


� Example: 

� Others: 

SPARQL: Solution set modifiers 

– ORDER BY DESC(?x) 

• Arrange in descending order 

– LIMIT n 

• Include only first n solutions 

– OFFSET n 

PREFIX f: 

SELECT ?p ?a 

WHERE { ?p f:age ?a } 

ORDER BY ?a 

• Include solutions starting from index n+1 

– SELECT DISTINCT 

• Do not duplicate solutions 



SPARQL: Constructing graphs 

� Annotate people with age below 26 as young people 


30 

f:john 

25 

f:age 

f:age 

Data “Query” 

f:loves 

f:jane 

26 

f:mary 

24 

f:age 

f:age 

PREFIX f: 

PREFIX rdf: 

CONSTRUCT { ?p rdf:type f:YoungPerson } 

WHERE { ?p f:age ?a . FILTER (?a < 26) } 

rdf:type 

f:john 

Result 

f:YoungPerson 

rdf:type 

f:mary 



SPARQL: several sources (1) 

f:john f:age "25"^^xsd:integer . 

f:bill f:age "30"^^xsd:integer . 

f:mary f:age "24"^^xsd:integer . 

f:jane f:age "26"^^xsd:integer . 

f:john f:loves f:mary . 

f:bill f:loves f:jane . 

f:john f:hasFriend f:bill 

http://users.jyu.fi/~minagy/itks544/rdf/people.rdf 

j:teacher rdf:type j:EducationJob . 

j:seniorResearcher rdf:type j:ResearchJob . 

j:juniorResearcher rdf:type j:ResearchJob . 

j:professor rdf:type j:ResearchJob, j:EducationJob . 

http://users.jyu.fi/~minagy/itks544/rdf/jobs.rdf 

f:john e:worksAs j:teacher . 

f:mary e:worksAs j:seniorResearcher . 

f:jane e:worksAs j:juniorResearcher . 

f:bill e:worksAs j:professor . 

http://users.jyu.fi/~minagy/itks544/rdf/employment.rdf 

@prefix j: . 

@prefix e: . 

@prefix f: . 

@prefix rdf: . 

@prefix xsd: . 

Prefixes 

From now on prefixes will 

be omitted to save space 



SPARQL: several sources (2) 

� Show me all people who have a friend together with all the people that 

are younger than 25 

Query 

SELECT ?per ?ag ?job 

FROM 

FROM 

FROM 

WHERE { 

?per f:age ?ag . 

?per e:worksAs ?job . 

?job rdf:type j:EducationJob . 

} 

Result 

per ag job 

f:bill 30 j:professor 

f:john 25 j:teacher 



SPARQL: Update 

� Also called SPARUL or SPARQL/Update 

� From January 2012 W3C Working Draft 

– http://www.w3.org/TR/sparql11-update/ 

� Insert, update and delete functionality 

� Example (taken from W3C doc): 

PREFIX dc: 

PREFIX xsd: 

DELETE { ?book ?p ?v } 

WHERE { 

?book dc:date ?date . 

FILTER ( ?date > "1970-01-01T00:00:00-02:00"^^xsd:dateTime ) 

?book ?p ?v 

} 



� Official specification: 

More on SPARQL 

– http://www.w3.org/TR/rdf-sparql-query/ 

� SPARQL update 

– http://www.w3.org/TR/sparql11-update/ 

� Tutorials: 

– http://jena.sourceforge.net/ARQ/Tutorial/index.html 

� Validator: 

– http://www.w3.org/RDF/Validator/ 



Homework 

1. Download and start Sesame yourself 

2. Create some RDF file 

3. Import it to Sesame 

4. Try a few queries

Lecture 10: Storing and querying RDF data

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?