Implementing the DoD Business Mission Area (BMA) Vision June 26, 2012 Department of Defense Deedee Akeo, Senior Enterprise Architect, DCMO

Establishing the Foundation 1 Department of Defense

2 BMA Vision using Semantic Standards For Official Use Only 2 Query BEA directly: Acq Domain Vocabulary HR Domain Vocabulary Log Domain Vocabulary Fin Domain Vocabulary Real Prop Domain Vocabulary Business Enterprise Architecture: BEA Airman Sailor Svc Member position billet OUID dept User executes BP BP executes via BEA directly  Enterprise analytics  Compliance  IRB/portfolio management BP models uniformly described (GFMDI) (EDIPI) DoD EA SameAs DIEA Domain Vocabulary SameAs OMG Primitives Conformance class 2.0 Data described in RDF Relationship described in OWL W3C Open Standards Legend: DoD Authoritative Data Source 2 DM2 Department of Defense

3 A Standards-Based Approach Department of Defense

4 Establishing the Foundation Real Prop Domain Ontology Log Domain Ontology BEA Content BEA Ontology LRPs/Bus Rules Standard E2Es Standard Bus Process Standard Data Elements BPMN Ontology DM2 Ontology Metadata/Vocabularies converted to ontology Domain Ontology's extend BEA core ontology Components map to BEA and extended ontology's Supported by Equipping The Workforce (ETW) Practicum HR Domain Ontology Air Force NavyArmy Department of Defense

5 BMA Foundational Ontologies P2P H2R DM2 BEA Notional View of Foundational Ontologies BPMN 2.0 Department of Defense

6 Phase I DCMO Ontology Architecture(DM2 2.02, BPMN 2.0, BEA 9.0)  Ontologies created by Small Working Teams (Engineers, Ontologist, SMEs,…)  DM2 Working Team (included representatives from DoD CIO DM2 WG)  BPMN 2.0 Working Team; BEA Working Team Created DM2 v2.02 Ontology  Based on DM2 Logical and Conceptual Models  Replaced IDEAS constructs with OWL constructs Created BPMN 2.0 Ontology  Contains BPMN Analytic Conformance  BPMN 2.0 specifications used to create ontology  Plan to submit to OMG as an industry standard Created BPMN 2.0 to DM2 Mapping Ontology  Maps some BPMN concepts and relationships to DM2 OWL  Mapping achieved via OWL sub-classing and chaining axioms Created BEA Ontology  BEA Ov-6c uses BPMN 2.0 concepts and relationships directly  Other BEA views (e.g. OV2, OV5a, OV5b, SV1, Svc, CV-2) map to DM2 directly via sub-classing Foundational Ontologies Implementation Department of Defense DM2 BPMN-to-DM2BPMN 2.0 BEA (non- BPMN) BEA class replacements Query BEA on DM2 or BPMN terms Upper ontology to all architecture ontologies

7 Status of BEA Ontology Framework Department of Defense Release I Delivery BEA 9.0 data was migrated into DCMO Phase I Foundational Ontologies and tested with SPARQL queries A “BEA Ontologies” repository has been established on Forge.mil –Version of DM2 v2.02, BPMN 2.0, and BEA ontologies have been submitted to the “BEA Ontologies” directory –Ontology versioning is currently specified via owl:versionInfo tag –A “BEA Ontology Framework Usage Guide” and readme file are also provided on Forge.mil

DoD BEA/ DIEA Federation Pilot 8 Department of Defense

9 Semantic Architecture Federation –Enable disparate sources of architecture to be logically integrated and accessed as a virtual single store, using semantic technology Make Compliance Easier –Enable a single user perspective on disparate compliance data Pilot Objectives Department of Defense

10 Uses DM2 and semantic technology Demonstrates –Enterprise information search capability with DM2 –Utilizing OWL language and infrastructure to define information models (DM2, BEA, DIEA) and execute queries –Mapping “upper ontology” to “lower ontology's” Ontology Mapping BEA OntologyDIEA Ontology User Interface RDB OWL to RDB map DM2 Ontology e.g. “what possible information exists that is within the scope of the concept of a DM2 Activity” Department of Defense

DIEA BEA DM2/BEA/DIEA Ontology Mapping Department of Defense

12 Federation Demonstrated DM2 Ontology 1. DM2 2. DM2->BEA 3. BEA->DIEA BEA-DIEA physical links r2rml BEA Architecture Data (RDB) DIEA Architecture Data (RDB) Department of Defense

13 Demonstrated feasibility of federating architectures to support compliance –Mechanically it is feasible, semantics provides easier infrastructure for scoping and linking –Ontological “similarity” is necessary for meaningful federation “you can’t just connect stuff” – Humans must do this Demonstrated ontology based-distributed and federated query –SPARQL queries used to query repositories –Human and machine readable, easier to understand and execute Demonstrated use of DM2 encoded in OWL –DM2 concepts sub-classed to create BEA/DIEA concepts Establish semantic technology environment for more complex federation use-cases - Pieces available, compliance use case not fully tested Pilot Summary & Lessons Learned Department of Defense

14 Incorporate BPMN 2.0 OWL-DL as an addition to the OMG standard Develop and incorporate DM2 v2.03 OWL-DL in DoDAF v2.03, Volume 3 Build target BEA RDF deployment platform Continue to equip the workforce (ETW) Component implementation of semantic standards Leverage BMA Foundational Ontologies Next Steps Department of Defense

15 Department of Defense

16 DoD Federated/Net-centricity and DCMO Approach “Enterprise Architecture Federation Strategy” on Semantic Alignment excerpt: A key goal of net-centricity is to enable semantic understanding of data so that interoperability can be achieved between any applications that have the ability to access and interpret the structural and semantic rules associated with data. DCMO- DoD Federated/Net-centric approach through semantic specifications: Use of semantic standards to realize DoD federated understanding of data o OWL (Web Ontology Language) o OWL-DL (Descriptive Logic) o SPARQL 1.1 (OWL Query Language) o BPMN2.0/BPMN 2.0 primitives (Business Process) o Standards adopted by DoD in the DISR A standards based semantic understanding between enterprise applications supports the shift from a stale data warehousing approach to federated dynamic retrieval of authoritative data sources approach “Net-centric Data Strategy” on Interoperable excerpt: Data Interoperability - The ability to share information among components while preserving its accuracy, integrity and appropriate use. The ability of two or more systems or components to exchange information and to use the information that has been exchanged. Goal: Decentralize data management to communities of interest (COIs) to allow prioritization and collaboration based on immediate operational needs while providing enterprise infrastructure for self-synchronization on a larger scale Department of Defense

17 Objectives: –Provide a common way to express architectures for DoD and other US NATO partners –Provide a common means for sharing architectural data Existing DoDAF DM Model –DM2 model is based on IDEAS architectural framework –Conceptual Data Model(CDM): Captures high level architecture concepts and relationships –Logical Data Model(LDM): Captures architecture detailed relationships on additional domains (Security, Locations, Process, etc.) –Physical Exchange Specification(PES): Provides a schema and approach to exchange DM2 data across the enterprise in DM2 representation *DoDAF Conformance is achieved when: –The data in a described architecture is defined according to the DM2 concepts, associations, and attributes –The architectural data is capable of transfer in accordance with the PES *As defined in: DoD Architecture Framework Version 2.0, Volume 1: Introduction, Overview, and Concepts Manager’s Guide DoDAF and the DM2 Department of Defense

18 DoDAF Conceptual Model Department of Defense

19 DoDAF DM2 Logical Model Department of Defense

20 DCMO DM2 OWL-DL Approach Department of Defense The goal was to describe DM2 completely in OWL open specification based constructs –Started with classes from the DM2 conceptual model –Expanded conceptual class model using details from DM2 logical model –All DM2 IDEAS based constructs that could be expressed in OWL were replaced with standard OWL constructs; e.g. o ideas:Thing => owl:Thing o Ideas:Name => rdfs:label o ideas:Type => owl:Class –All DM2 IDEAS based constructs that were not required to express BEA were not included

21 DM2 Ontology Sample Snapshot Department of Defense

22 Created DM2 Ontology  Based on DM2 Logical and Conceptual Model  Replaced IDEAS constructs with OWL constructs Created BMPN 2.0 Ontology  Use BPMN 2.0 specifications as guidelines to create ontology  Ontology closely resembles specifications  Flesh out BPMN Ontology with Signavio generated XML data  *Plan to programmatically migrate BPMN xml generated from tool, into “BPMN RDF” Created BPMN Ontology Mappings (Rules)  Mapped appropriate BPMN 2.0 classes to DM2 classes (sub-classing)  Mapped appropriate DM2 properties to BPMN properties (rules & chaining axioms) Transformed BEA BPMN related data (SPIN & SPARQLMotion)  BEA BPMN data: (Ov-6c, *E2E)  Replaced BEA BPMN related concepts with BPMN 2.0 Ontology concepts Transformed BEA non-BPMN data (SPIN & SPARQLMotion)  Non-BPMN data: (OV2, OV5a, OV5b, SV1, Svc, CV-2) to BEA Core Ontology  Mapped concepts to DM2 directly or indirectly(through non-BPMN ontology) Semantic BEA Approach Department of Defense

23 OWL is an industry standard to express Ontologies: –WC3 open specification –Support from Open Source and COTS tool vendors –Tools exist to create/visualize OWL files, and process OWL rules –Active growing community continues to submit new, rich, supporting specifications/capabilities; i.e. R2ML, RIF, SPIN, RDFa, etc. Contains a rich set of constructs and data types –Is extendable and provides a path for modular development & reuse –Captures both data, and rules that can be quickly adjusted in a controlled fashion vs. capturing rules in code with a long change/deploy process –Formal logic and supports reasoning Interoperability: –OWL files are easily shared, making structures visible and well understood –SPARQL query specifications provide solution to dynamic federated queries –Same concepts with different terms across federated ontologies can be resolved via OWL “sameAs” constructs Advances to goal of establishing DoD architectures that are people readable, machine readable and executable Benefits of OWL based Ontologies? Department of Defense

24 OWL2 (OWL-DL)  Why OWL2 over OWL1? o Increased expressiveness o property chains, disjoint properties, etc. o extended datatypes and data ranges o enhanced annotation capabilities  Why OWL Description Logic(DL) o Maximum expressiveness o Reasoning with completeness and decidability SPARQL 1.1  Why SPARQL 1.1? o Useful new features; e.g. “Aggregate” Semantic Standards Department of Defense