Inferray

Inferray is cache-friendly forward-chaining reasoner that supports the following rule sets :

• RDFS Full
• RDFS Default
• RhoDF
• RDFSPlus

Citation

When using Inferray, please use the following citation :

Julien Subercaze, Christophe Gravier, Jules Chevalier, Frédérique Laforest:
Inferray: fast in-memory RDF inference. PVLDB 9(6): 468-479 (2016)

Bibtex :

 @article{DBLP:journals/pvldb/SubercazeGCL16,
 author    = {Julien Subercaze and
           Christophe Gravier and
           Jules Chevalier and
           Fr{\'{e}}d{\'{e}}rique Laforest},
 title     = {Inferray: fast in-memory {RDF} inference},
 journal   = {{PVLDB}},
 volume    = {9},
 number    = {6},
 pages     = {468--479},
 year      = {2016},
 url       = {http://www.vldb.org/pvldb/vol9/p468-subercaze.pdf}
}

Inferray can be used either as a standalone reasoner (project inferray-core) or as a reasoner behind Jena (inferray bindings).

Sample Use

Standalone

Instantiate the reasoner, load an ontology and process the inference.

Inferray inferray = new Inferray();
inferray.parse(myOntology);
inferray.process();

Export as a Jena Model :

ExportUtils.exportToJenaModel(inferray);

Jena Reasoner

OntModel model = ModelFactory.createOntologyModel(InferrayReasonerFactory.INFERRAY_RDFSPLUS);
	... make some changes ...
	model.rebind();

Then use the OntModel as usual with Jena. Inferray performs incremental inference when new triples are added and restart a full inference whenever triples are deleted from the Model.

Configuration

You can configure advanced parameters of Inferray using the Configuration Builder. Parameters are self explanatory and listed in the following snippet :

final ConfigurationBuilder builder = new ConfigurationBuilder();
final PropertyConfiguration config = builder
				.setDumpFileOnExit(false)
				.setMultithread(true).setThreadpoolSize(8)
				.setFastClosure(true)
				.setRulesProfile(SupportedProfile.RDFSPLUS)
				.build();
final Inferray inferray = new Inferray(config);

Maven dependencies

First you need to setup the server in your pom.xml :

<repositories>
  <repository>
    <id>inferray-mvn-repo</id>
    <url>https://raw.github.com/jsubercaze/inferray/mvn-repo/</url>
    <snapshots>
        <enabled>true</enabled>
        <updatePolicy>always</updatePolicy>
    </snapshots>
  </repository>
</repositories>

Then use the following dependency :

<dependency>
  <groupId>fr.ujm.tse.lt2c.satin</groupId>
  <artifactId>inferray</artifactId>
  <version>0.0.2-SNAPSHOT</version>
</dependency>

To load only the core engine :

<dependency>
  <groupId>fr.ujm.tse.lt2c.satin</groupId>
  <artifactId>inferray-core</artifactId>
  <version>0.0.2-SNAPSHOT</version>
</dependency>

Correctness

You can check the correctness of Inferray, using Jena as ground truth using the inferray-correctness project.