1. © 2006, The MITRE Corporation Toward a Standard Rule Language for Semantic Enterprise Integration Ms. Suzette Stoutenburg 719-572-8323 suzette@mitre.org MITRE Sponsored Research © 2006, The MITRE Corporation Dr. Leo Obrst 703-983-6770 lobrst@mitre.org

2. © 2006, The MITRE Corporation Problem n To defeat emerging threats, C2 systems must be dynamic and adaptable n Separation of rules from application code supports the ability to dynamically modify system behavior in complex, changing environments n A rule language standard is required to enable agility and interoperability across the enterprise

3. © 2006, The MITRE Corporation Background Semantic Web Stack (partial) n Semantic Web standards provide infrastructure for specifying and operating over semantically meaningful information n Sets the stage for net-centric autonomous reasoners and M2M capabilities

4. © 2006, The MITRE Corporation Objective n Develop demonstrable recommendations for a standard rule language n Identify sponsor-specific requirements for rules and ensure those are captured in the standard n Build experiments to show how ontology and rule layers should interact and compare approaches n Look ahead to future requirements for the evolving standard

5. © 2006, The MITRE Corporation Activities n Examine use of rules to build dynamic net-centric services – Show how rules of engagement can be modified in real-time based on dynamic events – Extended first-year experiment using coalition battlefield scenario n Optimize the Logic Programming environment that integrates rules and ontologies

6. © 2006, The MITRE Corporation Prototype Architecture Events (GMTI, INTSUM) Situational Awareness Service Enterprise Service Bus 2. Message injection starts AMZI Logic Server 1 AMZI Logic Server 2 1. Logic Servers instantiated with Rule Sets 3. Adapter translates schema 4. New messages applied to Knowledge Bases Google Earth Client Adapter 6. SA Svc detects events on ESB, queries for alerts 7. Rules fire, produce “Single Integrated Picture” 8.Battlefield movement, Intel, messages and alerts displayed on client Message Injector 5. Rules fire to relate messages to Theater Objects Simulated Position of Blue Force Optimized Knowledge Base 1 Optimized Knowledge Base 2

7. © 2006, The MITRE Corporation destination Application convoy enemy sniper friend or foe? RECOMMENDATION: Take alternate route primary route alternate route ALERT: Intelligence report of enemy sniper in the vicinity Command and control of convoy movement in unsecured territory with situational awareness and rules of engagement

8. © 2006, The MITRE Corporation Ontology Overview TheaterObject ConvoyRegionOfInterestConvoyRouteConditionsAndAlerts

9. © 2006, The MITRE Corporation System Design Translation to Prolog Extensionalization User Queries Answers Changes Extensional Program (Prolog) Queries Answers Extensional Program (Prolog) Dynamic General Rules (Prolog) Intensional Program (Prolog) Intensional Program (Prolog) Developer Cardinality Rules (Prolog) rules ontologies SWORIER (Semantic Web Ontologies and Rules for Interoperability with Efficient Reasoning)

10. © 2006, The MITRE Corporation SWORIER Challenges KEY Fully implemented Partially implemented Not yet implemented n Logical Negation n Open/Closed World n Complementary/Disjoint Classes n Disjunctive Conclusions n Equivalence n Cardinality Restrictions n Enumerated Classes n Error Messages n Existential Quantification n Cyclic Hierarchies n Anonymous Classes n Duplicate Facts

11. © 2006, The MITRE Corporation SWORIER Efficiency Two Dynamic Changes (Online) Report the location and speed of a convoy Report the location and speed of an enemy motorized infantry unit Desired Time: Baseline: With Code Minimization: a few seconds 25.2 minutes 10 milliseconds Two Queries (Online) What are the locations and speeds of all known units? What are all of the alerts and recommendations? Desired Time: Baseline: With extensionalization: & Dynamic Changes: & Code Minimization: a few seconds 33 minutes 110 milliseconds 58 minutes 130 milliseconds Extensionalization (Offline) Desired Time: Baseline: With Avoiding Reanalysis: & Code Minimization: a few hours CRASH 13 hours 6.5 hours

12. © 2006, The MITRE Corporation Impacts n Advances critical sponsor needs – Agility in a changing world – Semantic interoperability and integration in a net-centric enterprise – Demonstration of dynamic service-oriented architecture n Contributes to the sound evolution of the rule layer of the Semantic Web – Recommendations for how language should evolve to support critical sponsor needs – Orchestration strategy for rules, ontologies, engines

13. © 2006, The MITRE Corporation Future Plans