Jena 2 - PowerPoint PPT Presentation

1 / 8
About This Presentation
Title:

Jena 2

Description:

[likes1: (?P pubs:servesBeer beertypes:Guinness) - (http://www.csd.abdn.ac.uk ... Free as in beer & speech, excellent community support and documentation. Questions? ... – PowerPoint PPT presentation

Number of Views:65
Avg rating:3.0/5.0
Slides: 9
Provided by: csdAb
Category:
Tags: jena

less

Transcript and Presenter's Notes

Title: Jena 2


1
Jena 2
  • http//jena.sourceforge.net
  • Gunnar AA. Grimnes

2
About Jena 2
  • Jena is a Semantic Web toolkit
  • More specifically a Java API for handling RDF.
  • Developed at HP Research Labs.
  • Now open source
    http//jena.sourceforge.net
  • Jena supports
  • RDF/XML, OWL, DAMLOIL, RDQL, N3, XMLSchema
    datatyping, database persistence.
  • Inference

3
Inference in Jena 2
  • Everything is pluggable.
  • Supplied reasoners are rule based.
  • Built-in rule-sets for OWL (Lite), DAMLOIL and
    RDFS.
  • Any reasoner could be made to work with Jena
  • Prolog, SHOQ(D)
  • InfModel is a (java)subClass of Model.

Ontology Model
Instance Model
Rules
Owl rules Daml rules RDFS rules
Reasoner
Hybrid Reasoner
Inferred Model
Backward chainer
Forward Chainer
4
Jena Inference
  • Not Complete more importantly, not sound.
  • Limited OWL support
  • Does not support oneOf, unionOf or complementOf.
  • Does not automatically fetch imported ontologies.
  • Not very efficient
  • Inference over winefood example ontologies is
    possible, but takes 800mb of memory.

5
Jena inference code example
  • ...
  • import com.hp.hpl.jena.reasoner.
  • import com.hp.hpl.jena.reasoner.rulesys.
  • ...
  • Model peoplemodelModelFactory.createDefaultModel(
    )
  • peoplemodel.read("people.rdfs")
  • Model basemodelModelFactory.createDefaultModel()
  • basemodel.read("department.rdf")
  • Reasoner reasonerOWLFBRuleReasonerFactory.theInst
    ance().
  • create(null).bindSchema(peoplemodel)
  • InfModel modelModelFactory.createInfModel(reasone
    r,basemodel)
  • Resource alunmodel.createResource(department.rd
    fapreece")
  • StmtIterator simodel.listStatements(alun,null,(RD
    FNode)null)
  • while (si.hasNext())

6
Custom rules in Jena
  • likes1 (?P pubsservesBeer
    beertypesGuinness) -gt (http//www.csd.abdn.ac.uk/
    ggrimnes preflikes ?P)
  • or
  • rule0 (?B foafknows ?A),
  • (?B foafknows ?X),
  • (?X foafmbox mailtoterry_at_acm.org),
  • (?P foafmbox mailtoapreece_at_csd.abdn.ac.uk) -gt
  • (?P foafinferred_knows ?A)

7
Conclusions
  • Jena has different goals than Cerebra.
  • A great general SW toolkit.
  • Still early days for inference framework (Beta in
    June03).
  • Free as in beer speech, excellent community
    support and documentation.

8
Questions?
  • ?

?
Write a Comment
User Comments (0)
About PowerShow.com