Empire: JPA for RDF & SPARQL
- 1. Empire: RDF & SPARQL Meet Java & JPA 2010 Semantic Technology Conference Michael Grove Director of Software Development Clark & Parsia, LLC. mike@clarkparsia.com http://clarkparsia.com -- http://www.twitter.com/candp
- 2. Background RDF & SPARQL You know what these are... Java Persistence API (JPA) Standard ORM API for Java developers Provides a way to map Java beans to relational data Part of Java SE and EE editions Implementations: Hibernate and Toplink. Empire Implements JPA for RDF databases Uses SPARQL rather than JPQL We're considering adding JPQL support Apache 2.0 License -- released in Feb 2010.
- 3. Why? Empire bridges the gap between traditional web apps & the semantic web JPA helps abstract away (most) of the RDF details Helps lower learning curve for new developers API, database, & query language independence Application code is written w/r/t to JPA API Developers only need to learn the domain API Sesame/Jena/Other bindings are injected at runtime -- invisible to the developer Helps open up the Semantic Web toolkit Can integrate semweb apps, like Pelorus, with traditional webapps Conversely, can integrate traditional Java web frameworks with the Semantic Web
- 4. Why do this... Graph aGraph = new GraphImpl(); URI aBook = aGraph.getValueFactory().createURI("urn:oreilly.com:product:9780596514129.IP"); aGraph.add(aBook, aGraph.getValueFactory().createURI("http://purl.org/dc/terms/publisher"), aGraph.getValueFactory().createLiteral("O'Reilly Media / Pogue Press")); aGraph.add(aBook, aGraph.getValueFactory().createURI("http://purl.org/dc/terms/title"), aGraph.getValueFactory().createLiteral("Switching to the Mac: The Missing Manual, Leopard Edition")); // ... setting the additional properties here ... aGraph.add(aBook, RDF.TYPE, aGraph.getValueFactory().createURI("http://vocab.org/frbr/core#Expression"));
- 5. When you can do this. Book aBook = new Book(); aBook.setTitle("Switching to the Mac: The Missing Manual, Leopard Edition"); aBook.setPublisher("O'Reilly Media / Pogue Press"); aBook.setIssueDate("2008-02-26"); // And so on... This code is... More clear More concise Easier to understand and maintain API independent
- 6. Advantages Rapid prototypes Lowers the learning curve by simplifying the codebase No more triples and graphs littering your code Utilities can generate your domain from your ontology Lets you concentrate on the more important things API independence Need to switch from Sesame Memory Store to TDB? Change one line of code and you're done. We standardize on the standards, RDF & SPARQL (and JPA), not a particular API like Sesame or Jena. Plays nicely with others... Can work alongside normal JPA implementations Provides a straightforward migration to a more semantic-web based infrastructure
- 7. Obligatory Technical Fu Provides support for Jena, Sesame & 4Store This includes TDB, SDB, Sesame Native, etc. Support planned for Mulgara, Bigdata, Virtuoso and others. Dialect implementations for both SPARQL and SeRQL. No complicated configuration and mappings Mappings use standard JPA annotations Configuration are basic Java properties files. Google Guice used to manage plugin architecture Javassist for bytecode enhancement
- 8. In case you're still awake... Really? JPA? All of it? Not quite... We don't support things that are obviously relational, such as @Table or @Column. We also don't support parts such as @SqlSetResultMapping or JPQL, but we're planning on doing that Real Soon Now And we're still a little fuzzy on transactions and locking Most RDF stores don't neatly support these, or do so inconsistently We provide a naive approach, but that too will improve
- 9. Questions?
- 10. For more information... A detailed overview: http://semanticuniverse.com/articles-empire- rdf-sparql-meet-jpa.html The WIKI: http://wiki.github.com/clarkparsia/Empire/ The mailing list: http://groups.google.com/group/empire-rdf