Your Federal Quarterly Tax Payments are due April 15th Get Help Now >>

userSmarts® The Semantic Web Middleware by fjn47816

VIEWS: 10 PAGES: 33

									       userSmarts®
The Semantic Web Middleware
                                                    From Research to Engineering



          Yaser Bishr, PhD
         Image Matters LLC.
         yaserb@imagem.cc
        FOSE 2004 March 24th

         Image M atters Copyright Proprietary Information.
                                                                         1
                         History

• Started in 2000 and                     • Achievements
  opened our doors as an                           – Image Matters has been
                                                     investing in next generation
  LLC in January 2001                                Smart Web Services
• Won SBIR Phase I and                               (Semantic Web); ~$1.5M.
  Phase II for Semantic                            – Well on track for Release
                                                     of our Smart Service
  Technologies                                       Middleware (userSmarts®)
• Currently in Phase III                             by the end of the year.
                                                   – We are the leading
                                                     Interoperability
                                                     Engineering resource for
                                                     the OpenGIS Consortium.

                  Image M atters Copyright Proprietary Information.
                                                                               2
             Business Activities

• Advanced Geospatial Semantic R&D
• userSmarts® Product Development
• Interoperability Engineering Consulting




                Image M atters Copyright Proprietary Information.
                                                                    3
                      Mission




To be the Leader in Products and Services for the
             “Geospatial Smart Web”




                Image M atters Copyright Proprietary Information.
                                                                    4
Characterizing the Problem: Decision Making


                         Orient




                         Situation
     Observe                                                       Decide
                        Awareness




                              Act

               Image M atters Copyright Proprietary Information.
                                                                            5
                      Setting The Stage

• The Semantic Web
   – An extension of the current web in which information is given well-
     defined meaning, better enabling computers and people to work in
     cooperation. (W3C)

• Situation Awareness
   – The perception of the elements in the environment within a volume of time
     and space, the comprehension of their meaning, and the projection of their
     status in the near future. (Endsley, 1988).

• Components
   –   Sensory Network to sense the environment  time, space, behavior, state
   –   Knowledge Model of the environment  Comprehend the environment
   –   Reasoning and Analytical Tools  Projection and Status
   –   Usability Engineering and HCI  Visualize and Convey the situation

                            Image M atters Copyright Proprietary Information.
                                                                                6
   Situation Awareness Problem Statement
• Present systems                                               • Situation Awareness
   – Information poorly organized,                                        –      Actors
     unfiltered, unprocessed                                              –      Goals and objectives
   – Information difficult to search/access                               –      Courses of action
   – Information has marginal relevance to                                –      Tactics
     decision-makers                                                      –      Critical events
                                                                          –      Space, Time and Flow
• Decision-makers want
   – Reduce the OODA time Loop
   – Shared understanding among                                                           Orient

     stakeholders
   – Near real time Common Relevant
     Operational Picture (CROP)
   – Actionable knowledge                                               Observe                         Decide

   – Tailored decision frameworks


                                                                                            Act


                             Image M atters Copyright Proprietary Information.
                                                                                                          7
       Situation Awareness System Concept

• Dynamically collects and organizes information that
  is relevant to a situation and the commander’s
  decision process, i.e., Information in “Context”
   – Goal driven (not just data driven)
   – Complex operations with multiple stakeholders
• Supports/integrates
   –   Warnings and Notifications: Support Push/Pull instead of only Pull
   –   Dispatch (CAD) and Vehicle Location (AVL) technology
   –   Damage Assessment
   –   Command & Control
   –   Interagency Coordination
   –   Public Information and Warning



                              Image M atters Copyright Proprietary Information.
                                                                                  8
                          Semantic Model

• Model - Set of quadruples
• Quadruple – truth assertion
   – <Context, subject, predicate, object>
• Class – resource classifier
   – subclassOf
• Property – predicate types                                Context
   – Domain constraints                                                      Resource
   – Range constraints                                                                         Property


• Schema – vocabulary of                                       Subject
                                                                                                          Value

  Types and Predicates                                                             Predicate


   – Defines a closed KB namespace                                                               Object

   – Can import other schemas
• Containers
   – Collections of Values
        • Resource
        • Literal              Image M atters Copyright Proprietary Information.
                                                                                                                  9
Top Level Ontology in UserSmarts




          Image M atters Copyright Proprietary Information.
                                                              10
      Scalable Semantic Interoperability
                             Database

                                    S-O                                            App
      App                           Map
               S-O                                                          S-O
               Map                                                          Map

                            Ontology
                                                                            S-O
                                                                            Map
                 S-O
                 Map                                                              Database
                                           S-O
                                           Map

    Database
                                             App

Explicit semantics =  Interopability/Reuse/Utility
                        Image M atters Copyright Proprietary Information.
                                                                                             11
Person: IRA-1

        Personal:
                                       Related Resources
        First: Abdul
        Last: Bahieri
                                                                                                               ORG
        Gender: Male                      EQP
        DOB: 2003-36=1967
                                                                                                            (UoDetroit)
                                         (SUV2)        operates                             attended
        POB: Iran
        Nationality: US                                                     PER
        Id: (US-passport#)                                                (IRA-1)
        Height:
        Weight:
                                                   residesAt                                   bizPartner
                                         PLA
        Complexion:                                                                                             PER
                                      (Ashburn)
        Hair Color:                                      attends                                              (SPA-1)
        Eye Color:                                                              owns
       Contact:
       Other Name:
       Alias:
                                                                                            ORG
       Email:
       Work Phone:
                                                     ORG                                 (LOUGAS)
       Home Phone:                             (Khalid Rasheed
       Mobile Phone:                            Khalid Mosque)

         Address:
         Home Address:         Personal Narrative:
         Home City: Ashburn
                               Male, age 36, no wants/warrants/aliases, one speeding traffic ticket issued by Loudoun County in 1998; Born
         Home State:Virginia
                               Iranian; immigrated to the US in 1987 via London, England. He is a legalized US citizen and maintains a valid
         Home Zip:
                               US Passport. Established initial residence in Detroit, Michigan; in 1992 graduated from the University of Detroit
         Home Country:
                               with a degree in Industrial Engineering; relocated to the Washington DC area in 1993. In 1994, established the
         Work Address:
                               Loudoun Industrial Gas Supply Company (LOUGAS) in Leesburg, Virginia. He is the principal owner/operator
         Work City:
                               and his business partner is known to be Franco Luis Marico. He is divorced from Soulham Kali Bahieri who
         Work State:
                               returned to Tehran in 1996; they have no children; he lives in Ashburn, Virginia; he attends the Khalid Rasheed
         Work Zip:
                               Khalid mosque in Great Falls, Virginia. His extended family is in Tehran, Iran; there are no reported concerns
         Work Country:
                               expressed by the Government of Iran. He is known to be critical of US Mid-East policy in general.
                               Importantly, he is highly critical of the expanded US anti-terrorist efforts implemented after September
 Attachments:                  11,2001.
                                                     Image M atters Copyright Proprietary Information.
                                                                                                                                      12
Person: ALF-2
                                               Related Resources
      Personal:
      First: Pang
      Last: Mango
      Gender: Male
      DOB: 2003-30=1973                EQP
      POB:                            (TRK1)         operates                   PER
      Nationality: Indonesia                                                  (ALF-2)
      Id: (Indonesia-passport#)
      Height::
      Weight:
      Complexion:
      Hair Color:                                                 residesAt
      Eye Color:                                                                           employeeOf
      Contact:
      Other Name:
      Alias:
                                                                    PLA                                        ORG
                                                                  (OR, VA)                       ORG
      Email:                                                                                                  (WSA)
      Work Phone:
                                                                                                (SAS)
      Home Phone:
      Mobile Phone:               Personal Narrative:
                                  Pang Mango (Alfi-2)is 30. It was noted that when in the Indonesian Security Service, Mango was qualified as a
      Address:                    Munitions Specialist
      Home Address: unknown
      Home City:
      Home State:
      Home Zip:
      Home Country:
      Work Address: SAS, Front Royal Airport
      Work City: Front Royal Airport
      Work State: VA
      Work Zip:
      Work Country: USA

 Attachments:
                                                        Image M atters Copyright Proprietary Information.
                                                                                                                                             13
                         Knowledge Base
• Semantic Repository                                 • Implementations
   – Type & Property Schemas                                   – Oracle 9i Version
   – Contextual Knowledge Models                                         • Native XML types
                                                                         • Spatial Data Option
• Folder Service                                                         • Oracle Advanced Queuing
   –   General Resource Repository                                                – XML payload
   –   Location Organized Folders                                                 – PL/SQL Filtering
   –   Mobile Objects                                          – Berkley DB
   –   Situation Awareness Folders                             – Planning to implement on
• Query Services                                                 Oracle 10G
   – Semantic
   – Spatial
                                             Mobile Objects
   – Temporal
                                                                         Model
• Notification                                                                                        Message Queue
                                                                                                       Message Queue

   – Knowledge Models                     <S, P, O>                                                        Message Queue
                                          <S, P, O>                      Model       Knowledge Base
   – Folders & Resources                  <S, P, O>

                                                    Spatial Resources
                              Image M atters Copyright Proprietary Information.
                                                                                                              14
                    UserSmarts®

• Context-Sensitive™ Personalized Service delivery
• Enterprise Architecture based on J2EE
• Middleware Technology
• Real-time Event Driven
• Agent Framework for Goal Oriented Programming
• Information Model based on Ontology (RDF+OWL)
• Knowledge Representation Framework OFC™ (Ontology
  Foundation Classes)
• Highly Scalable Knowledge Persistence System
• Inference Capabilities including Horn Clause and Event-
  Condition-Action (ECA)


                     Image M atters Copyright Proprietary Information.
                                                                         15
                             Messaging        Calendar         Directory         Identity
                              Server           Server           Server            Server
                                                                                                                  EIS
                              Instant                           Secure
                                                Portal                            Mobile
                             Messaging                          Remote
                                                Server                            Access
                              Server                            Access
                             Portal Services
                                                                                                             Geodata

Workstation                   Public Access
                 Load                                                   App
                                 Portal                                                      Registry
              Distribution                                            Services
  Thick                        Emergency
  Client       Trusted
                                Planning                                                     Gateway
               System                                                      KB
                              Center Portal                                                   Service
   Web          DMZ
 Services      Routers
                              EOC Portal
                                                                       Info                  Context       Sensors
                                                                      Delivery               Agents
                Proxy
  Thin
               Servers
  Client

 Mobile
                                                                      userSmarts
 WIreless                                                      J2EE Platform                            Messages &
  Client
                                                                                                          Alerts
                             Storage,
  User                       Clustering
                                                               Enterprise
                                                               clustering
                                                                                   Storage
                                                                                  Solutions             Distributed
                                                               Solutions
  Layer       DMZ
                             Systems
                                                                                                        Resource Layer
                                            Application Layer
                                         Image M atters Copyright Proprietary Information.
                                                                                                             16
Image M atters Copyright Proprietary Information.
                                                    17
Image M atters Copyright Proprietary Information.
                                                    18
Image M atters Copyright Proprietary Information.
                                                    19
Image M atters Copyright Proprietary Information.
                                                    20
Image M atters Copyright Proprietary Information.
                                                    21
                Final Remarks

• Geospatial Semantic Web is a key technology for
  Situation Awareness
• Well defined Geospatial Knowledge Model
  (Ontology + Rules) is key to Situation Awareness
• CROP is enabled by providing Context Sensitive™
  Actionable Information to users
• userSmarts is a technology that enables highly
  scalable and agile active/reactive Situation
  Awareness Applications

                  Image M atters Copyright Proprietary Information.
                                                                      22
              Major Challenges

• Building “consistent” Ontology
• Management of Knowledge
• Extending Knowledge while maintaining
  consistency
• Integrating Ontologies with existing ES
• Trust
• Change of ideology: From modeling
  structures to modeling logical relationships
• Diffusion of Innovation
                 Image M atters Copyright Proprietary Information.
                                                                     23
  Thank You


yaserb@imagem.cc




  Image M atters Copyright Proprietary Information.
                                                      24
Back up Slides




  Image M atters Copyright Proprietary Information.
                                                      25
              Geospatial Knowledge Layers



   Modeling        Run Time                                    Modeling               Run Time
    Tools           Agents                                      Tools                  Agents
       Inference Engine                                                        Inference Engine
                                                       (Road Crosses River) and (Road is Above
  Domain            Domain
                                                       River) then (Road is Bridge). (Road Below
 Knowledge         Knowledge
                                                              River) then (Road is Tunnel).
                                                        X Adjacent to Y Y Adjacent to X. X
                                                       Above YY Below X. X Entered Y  (X
   Core Geospatial Knowledge
                                                      Inside Y) and (Y = Polygon) and (Y Contains
                                                                           X)

Feature+ Geometry + Topology +
                                                                                    GML
           Temporal


      Abstract Model                                    Implementation Example

                               Image M atters Copyright Proprietary Information.
                                                                                                  26
          Geospatial Knowledge Layers



   Modeling        Run Time                                     Modeling                  Run Time
    Tools           Agents                                       Tools                     Agents
       Inference Engine                                                            Inference Engine
                                                         (Road Crosses River) and (Road is Above
  Domain            Domain
                                                         River) then (Road is Bridge). (Road Below
 Knowledge         Knowledge
                                                                River) then (Road is Tunnel).
                                                          X Adjacent to Y Y Adjacent to X. X
                                                         Above YY Below X. X Entered Y  (X
   Core Geospatial Knowledge
                                                        Inside Y) and (Y = Polygon) and (Y Contains
                                                                             X)

Feature+ Geometry + Topology +
                                                                                         GML
           Temporal




                               Image M atters Copyright Proprietary Information.
                                                                                                      27
      Example of Semantic Modeling Tools
          Modeling
           Tools
                        Inference Engine
            (Road Cross River) and (Road is Above
           River) then (Road is Bridge). (Road Below
                  River) then (Road is Tunnel).
                                                                                      Tools to Construct
            X Adjacent to Y Y Adjacent to X. X                                       and Semantically
           Above YY Below X. X Entered Y  (X                                        Validate These
          Inside Y) and (Y = Polygon) and (Y Contain                                  Layers
                               X)
                              GML


        Transportation
                                                                                     US State
                                                                                  Inside
                            Cross Waterway
Air              Road                                                 Cross
                                      Over
                                                                          National Park
                                              Under
        Bridge             Tunnel
                                                                                 geometry: Polygon?
                             Image M atters Copyright Proprietary Information.
                                                                                                      28
                         S-O Editor
•Visualize the
structure of both
schema and ontology   Database Schema for                           Ontology for
                                                                      Facilities
• Assign explicit
                      Facilities Management
                                                                   Infrastructure
semantics to schema
elements
• Validate logical
consistency of
mappings
• Highlight
recommended
mappings
                                                           S-O Maps
• Supports standard
encodings of schema
and ontology
                        Image M atters Copyright Proprietary Information.
                                                                                    29
         Geospatial Knowledge Layers



   Modeling       Run Time                                    Modeling                Run Time
    Tools          Agents                                      Tools                   Agents
       Inference Engine                                                        Inference Engine
                                                         (Road Cross River) and (Road is Above
  Domain            Domain
                                                        River) then (Road is Bridge). (Road below
 Knowledge         Knowledge
                                                               River) then (Road is Tunnel).
                                                         X Adjacent to Y Y Adjacent to X. X
                                                        Above YY Below X. X Entered Y  (X
   Core Geospatial Knowledge
                                                       Inside Y) and (Y = Polygon) and (Y contain
                                                                            X)

Feature+ Geometry + Topology +
                                                               Standard Spatio-temporal model
           Temporal




                           Image M atters Copyright Proprietary Information.
                                                                                                  30
                         Example: With the Semantic Web
              NIMA Data
                                                                                                                                 Army Data
          FeatureID:      1234                            Query Answer: No
          Type:           Interstate                                                                     FeatureID:      2356
          Attribute List: name: I95                                                                      Type:           ArmoredTruck
                          length: 35                                                                     Attribute List: Licence: Arm234
                          pavmentType: Asphalt                                                                           CargoType: Toxic Chemical
                          speed limit: 65                                                                                Weight: 60 Tons
                          MaxWeight: 70 Tons                                                                             speed limit: 50
                          TruckPermission: 7PM – 6AM
                                                                                                                   Truck           CargoType
                         Road Network                                                                                                                SubPropertyOf

                                                                                  ArmoredCargoCarrier             CargoCarrier      DecontaminationTruck
    Secondary Road            Interstate         Street




                   UnionOf
Ontology                                                                                                                         HazmatCargo
                                                                                                                                                           HazmatType
            DriveOn
                                                                                                                                                    Domain
                  HazmatTruck

                                                                                                                                                   Chemical, Biological
       Domain                                                                 US DOT
                                                 FeatureID:        1234
                                                 Type:             Interstate
       Time Permission
                                                 Attribute List:   name: I95
                                                                                                                                         Ontology 2
                                                                   length: 35
                                                              Image HazmatPermission: 12AM-6AM
                      Range
                                                                    M atters Copyright Proprietary Information.
                                   Time Period                                                                                                      31
              What is the Technology Stack?
             Ontology            App.            App.          App.           Digital
                                                                             Signature
             Metadata
                               Domain           Domain       Domain             P3P
              (None)
                                                                             Key Mngmt
                      Core Geospatial Ontology (None)                        Web Trust

                    GML (OGC)              OWL (W3C) + RuleML                  (W3C)

             XML/S (W3C)                     RDF (W3C)
                   XML (W3C)                       Namespaces (W3C)
                      URI (W3C)                    Unicode (Unicode.org)

•   RDF provides the basic building blocks for constructing logical statements.
•   The OWL and RuleML layer enables writing ontologies and is built upon RDF and XML/S
•   The GML Layer provides the atomic spatio-temporal types as well as the generic
    feature model
•   The Core Geospatial Ontology Layer provides the high-level semantic rules and
    constraints for geospatial applications
•   The Ontology Metadata layer enables us to search for ontologies and evaluate their
    fitness for use
•   Web Trust, Digital Signature, and Key Management provide mechanisms for
    applications to determine trust level, and it also enables ontology providers to put
    certain access restrictions on all orMpart of the ontologies.
                                     Image atters Copyright Proprietary Information.
                                                                                         32
     NIMA Geospatial-Intelligence Knowledge Architecture

                                                          Core Geospatial                                               NTAD
        FIPS
                                      Order of Battle                                   Transportation


                                                               Ontology
                                                                                                                        NIMA
                                       Military
      Transp.                         Equipment                Engine                            Logistics

                                                                                                                            Place
                                                                                                                            Names

   Imagery                                      Knowledge Base APIs
                Utilities                                                                                               Equip
                                                                                                                        ment
 Secure Network

    Best Team.                                          Capability              C2           Tasking         Planning   Order of
      agent            Notification     Imagery                                agent          agent           agent      Battle
                         agent           agent            agent                                                          agent



FCS Knowledge
    Base
                            Army Future Combat System, and other customer systems
                                                  Image M atters Copyright Proprietary Information.
                                                                                                                            33

								
To top