SQUIRE: SPARQL Query Recommendations by Example

This is a library that, given a SPARQL query that can be satisfied by a certain RDF dataset, recommends semantically similar queries that can be satisfied by another dataset.

Building

You can build the project using Maven 3 and a Java 8 SDK.

First checkout this project.

git clone https://github.com/carloallocca/Squire.git

Enter the directory you just checked out.

cd Squire

From the project directory you can build all the modules by running

mvn install -Dmaven.test.skip=true

When all unit tests work offline, we will stop recommending the -Dmaven.test.skip flag.

If you want to build only some modules individually:

• squire : the core library, build this first.
• launcher : command-line Java application as executable JAR file, depends on squire
• websquire : Web Service version with HTTP API available as a WAR archive.

Running

After building the project:

• Command Line : execute the JAR file in launcher/target to get the command syntax.
• Web Service : drop the WAR file found in websquire/target into your Web container.

Command line

Enter the directory containing the JAR

cd launcher/target

Run the executable JAR once to get the list of available commands

java -jar org.mksmart.squire.squire-launcher-{version}.jar

(version can be something like 1.1-SNAPSHOT)

For example, to tell it to index two SPARQL endpoints:

java -jar org.mksmart.squire.squire-launcher-{version}.jar index \
http://opendatacommunities.org/sparql \
http://data.admin.ch/query/

After indexing them, use them to recommend the equivalent of a SPARQL query that works on the first endpoint (-s) to the target endpoint (-t):

java -jar org.mksmart.squire.squire-launcher-{version}.jar \
-s http://opendatacommunities.org/sparql \
-t http://data.admin.ch/query/ \
recommend "SELECT DISTINCT ?s ?p ?o WHERE {?s a <http://data.ordnancesurvey.co.uk/ontology/admingeo/District> . ?s ?p ?o }"

This process will generate a report after it ends. If you also want it to print a log to file while it executes, add the -l option.

Authors

• Carlo Allocca, Samsung Inc.
• Alessandro Adamou, NUI Galway
• Mathieu d'Aquin, NUI Galway

Credit goes to the Knowledge Media Institute of The Open University, UK, for kickstarting this project.

Permissions

You can use, fork and republish this software as you like so long as you retain these permissions and give us credit, preferably by citing this open access paper:

Carlo Allocca, Alessandro Adamou, Mathieu d'Aquin, Enrico Motta:
SPARQL Query Recommendations by Example. ESWC (Satellite Events) 2016: 128-133, 
https://doi.org/10.1007/978-3-319-47602-5_26