perspectives on the need for semantic technology
DESCRIPTION
Perspectives on the Need for Semantic Technology. James Milligan Information Directorate Air Force Research Laboratory [email protected]. Presentation Outline. Provide some examples of military problems and needs ripe for semantic technology - PowerPoint PPT PresentationTRANSCRIPT
1
Perspectives on the Need for Semantic
Technology
James Milligan
Information Directorate Air Force Research Laboratory
2
Presentation Outline
• Provide some examples of military problems and needs ripe for semantic technology
• Describe some AFRL activities making use of semantic technology as part of the solution
• Conclusions
3
Heterogeneous EnterpriseProblems
• Achieving and maintaining systems interoperability
– Incompatible data, data models, services, and applications.
• Challenges exist in sharing information across domains
– Insufficient information sharing capabilities that allow effective information exchange across multiple communities of interest, security enclaves, organizational boundaries, and infrastructures.
• Policy conflicts can hamper efficiency and effectiveness of joint and coalition operations
– Insufficient means to encode policies for semi-autonomous interpretation, negotiation, enactment, and enforcement.
4
Heterogeneous EnterpriseNeeds
• Systems that interoperate (even in the face of change)
– Compatible data, data models, services, and applications.
• Cross-domain information sharing
– Information sharing capabilities that allow effective information exchange across multiple communities of interest.
• Policy specification and enforcement
– Tools and mechanisms to encode policies for semi-autonomous interpretation, negotiation, enactment, and enforcement.
5
Policy Enforced Interoperable Communities of Interest
Theater Region (PACOM)Region (PACOM)
COIInformation
Space
COIInformation
Space
COIInformation
Space
Region (NORTHCOM)
COIInformation
Space
Info
6
Situational AwarenessProblems
• Maintaining situational awareness is difficult if the necessary collection assets cannot be rapidly deployed and connected
– Inability to autonomously manage and network theatre assets for rapid situational awareness.
• Correlating and integrating vast amounts intelligence data remains a hard problem
– Particularly as new sensor systems and platforms come on-line, it is difficult to effectively correlate, disambiguate, deconflict and combine sensor and human intelligence data into a common contextual model.
• Human interpretation of intelligence information is time consuming and sometimes error-prone
– Insufficient ability to rapidly interpret vast amounts of intelligence information.
7
Situational AwarenessNeeds
• Enabling situational awareness
– Ability to autonomously manage and network theatre assets for rapidly achieving situational awareness.
• Intelligence data fusion
– Ability to effectively correlate, disambiguate, and combine sensor and human intelligence data into a common context and information model.
• Rapid interpretation of intelligence
– Ability to rapidly interpret vast amounts of intelligence information.
8
Intelligence Collection and Analysis
Enable
Fuse
Interpret
9
Effects-Based OperationsProblems
• Ensuring that critical command and control systems continue to operate as needed
– Information systems require a high level of human intervention to keep them operational during mission preparation and execution.
• It is difficult to synchronize the application of diverse, distributed forces in time-critical situations
– Execution management capabilities are challenged in near real-time conditions in order to dominate an adversary and achieve the desired effects.
• Rapidly assessing that combat operations are achieving the desired effects is a challenge
– Real-time effects-based assessment of combat operations is problematic.
10
Effects-Based OperationsNeeds
• Command and control systems resource management
– Self-aware systems that can learn, adapt, and heal themselves.
• Rapid employment of agile forces
– Near real-time, dynamic synchronization of the employment of distributed forces to dominate an adversary and achieve desired effects.
• Real-time effects assessment
– Provisioning of real-time effects-based assessment of combat operations.
11
C2, Synchronization, Assessment
Targeting Decision
CAOC
Collaboration
NCES MessagingService
NCES CollaborationService
SubscribersCAOC
BDA
Notification
ConstellationNet
Semantically Tagged Info
NCES MessagingService
CAOC – Combined Air and Space Operations Center
NCES – Net-centric Enterprise Services
Intelligence Report
ENEMY
TANK
Weapons Effects AssessmentWeapons Effects Assessment
12
AFRL Semantic WebTechnologies & Applications
• Enabling Technologies
– Ontologies
– Knowledge Bases
– Artificial Intelligence
– Expert Systems
– Intelligent Agents
– Machine Inferencing, Reasoning and Learning
– Formal Methods
– W3C Standards, Protocols, and Reference Implementations
• Defense Applications
– Net-Centric Operations (e.g., NCES)
• Service Discovery, Composition, Mediation, Workflow Orchestration, and Execution
– Natural Language Processing
– Systems and Information Modeling, Integration, and Interoperability
– Knowledge Acquisition/Representation
– Cognitive Systems
– Modeling and Simulation
– Policy Representation and Enforcement
– Collaboration
– Cross-domain Information Sharing
– Resource Management
– Effects Based Operations
– Adversarial Modeling/Cyber Operations
– Decision Support/Planning/Predictive Environments (e.g., CPE)
– Multi-Platform/Source Intelligence Fusion
– Homeland Security
13
Representative AFRL Semantic Research and Applications
• Information Transformation
– Semantic Interoperability
• Anticipatory Environments
– Dynamic Air & Space Effects Based Assessment (DASEA)
• Collaboration
– Distributed enterprise object models
– Integrating Human Centeredness in the Design of Collaborative Systems
• Cross-Domain Information Access (CDIA)
– Foundation technologies
• Defensive Cyber Operations
– Cyber Situational Understanding (Event Pattern Ontologies SBIR)
• NETWAR: Strategic Attack Prediction and Detection
• Command and Control Resource Management System
• Link Discovery & Pattern Learning
• Heterogeneous Urban RSTA—Reconnaissance, Surveillance, Target Acquisition—Team (HURT)
• SIGINT Sensor Management
• Semantic-based Policy Enforcement
• Effects Based Operations
– Ontology for Scenario Generation SBIR
• NeXt Generation (XG) Program
– OWL Policy Language Development
• Command and Control Mobility Projects
– Intelligent Semantic NOTAM Query Capability
• Formal Methods
14
Representative AFRL Efforts Employing Semantic Technologies
• Broad Agency Announcements (BAAs)
– Semantic Interoperability
– Rapid Processing of Intelligence Data
– Commander’s Predictive Environment (CPE)
– Infospace Concept Exploration and Development
– Multi-Platform SIGINT Fusion and ISR Management
– Tangram: A Fully Automated, Continuously Operating, Intelligence Analysis Support System
– Intelligence Fusion for Targets-Under-Trees
• FY06 Small Business Innovative Research (SBIR) Program topics:
– AF06-047 Semantic Interoperability of C2 Tools and Technologies
– AF06-049 Real-Time Effects Assessment Management System
– AF06-052 Semantically Correct Interoperability of Executable Architectures
– AF06-053 Knowledge-based Technologies to Support Predictive Mission Awareness
– AF06-060 Enabling Monitoring and Analysis of Concept-Based Event Information in Text
– AF06-061 Multi-INT Ontology Mediation Services
– AF06-066 Systems-of-Systems Data Utilization Patterns
– AF06-077 Command Decision Support and Explanation from Fused Structured and Unstructured Information Sources
15
Conclusions
• Semantic technology holds great promise in addressing many of the problems and needs identified
• Additional investments are needed to mature the technology and make it easier to use and deploy in military applications
• An incremental approach toward widespread adoption seems likely – some semantic technologies will gain traction sooner than others
• We (academia, industry, government) need to do it together, and progress is rapidly being made from a historical perspective