Effective Ontology Development using UML and Enterprise Architect TDWG Technical Architecture Group – 17 Oct 2011 Sam Mancarella Chief Technology Officer Sparx Systems Pty Ltd Overview • Ontologies in Enterprise Architect – MDG Technology for ODM – OWL / RDF • Challenges of Ontology development – Collaboration, Maintenance, Reuse, Scale – How Enterprise Architect meets them …but first… an introduction to Sparx Systems and Enterprise Architect Sparx Systems • Established in 1998 • Leading provider of modeling tools – Support for Integrated Development Environments (IDEs) – Support for industry-standard modeling languages, technologies and frameworks (UML, SysML, BPMN, SoaML, TOGAF) • Representation in over 60 countries – Value-added resellers (solution providers) – Resellers, Training Partners – Sister companies (Japan, Central Europe, Sth America, NZ) • Headquartered in Creswick (central Victoria, Australia) Enterprise Architect • UML 2.3 powered, team-based modeling environment – Full lifecycle management – Business modeling, Systems engineering, Enterprise architecture, Requirements management, Software design – Scaleable, High-performance, Multiuser – Visual Execution Analysis, Simulation – Template-based reporting (RTF, HTML) – Code generation/roundtrip support for > 10 languages – Extensible, Customizable • Over 250,000 active licensees worldwide • Award Winning Enterprise Architect MDG Technology for ODM • Based on Ontology Definition Metamodel – OMG Standard (May 2009) V 1.0 – UML Profiles for RDF, OWL – Diagram Types, Construct Toolboxes, Import/Export Script MDG Technology for ODM 2 Diagram Type Toolbox Property Definition Subtyping Semantic Relationship MDG Technology for ODM 3 MDG Technology for ODM 4 MDG Technology for ODM 5 • Facts (ie: Things) – owlIndividual Instances of owlClasses Challenges of Ontology Definition • Scale • Ontologies can get really big! – 30,000+ artifacts in NCI Thesaurus • Difficult to Maintain • Need to Reuse Challenges – Scale 2 • What’s changed between revisions? Challenges – Scale 2 • Model Differencing (add, delete, change, move) Challenges – Scale 2 • Visual Differencing Challenges – Scale 3 • How can we reuse what we’ve already done again? Challenges – Scale 3 • Model Patterns (reusable model fragments) • Copy-paste between multiple instances Challenges of Ontology Definition • Teamwork • Multiple people develop ontologies – SMEs, Analysts, Practitioners, Stakeholders • All have a say in ontology development • Need to contribute value • Need to use the ontology Challenges – Teamwork 2 • Team Review (conversations about the model, in the model) Challenges – Teamwork 3 • Fully Customizable Environment Challenges – Teamwork 4 • Project Planning, Resourcing, Management Challenges – Teamwork 5 • Project Audit (who changed what, when?) Conclusion • MDG Technology for ODM: – – – – RDF & OWL support in Enterprise Architect owlClass, owlObjectProperty, owlIndividual Subclassing, Restriction (cardinality, value) Semantics, Facts • Ontology development challenges – Scale (reuse, baselines, model comparison, manage change) – Teamwork (team review, customize, audit, project management) • Enterprise Architect provides best-of-breed tools to: – Mitigate design complexity – Enable greater stakeholder enablement – Ensure design correctness/completeness/relevance Learn More? • Enterprise Architect 9.1 Trial http://www.sparxsystems.com/products/ea/trial.html • Reviewer’s Guide [whitepaper] http://sparxsystems.com/downloads/whitepapers/EAReviewersGuide.pdf • MDA Overview [whitepaper] http://www.sparxsystems.com/bin/MDA%20Tool.pdf • Deployment of EA [whitepaper] http://sparxsystems.com/downloads/whitepapers/EA_Deployment.pdf • Sparx Systems Community http://community.sparxsystems.com/ • SparxSystems YouTube Channel http://www.youtube.com/user/SparxSystems