Space Surveillance Ontology Captured in an XML Schema

1
Space Surveillance Ontology Captured in an XML
Schema

• Daniel L. Brandsma
• May 23, 2001

Organization D540 Project 03004000
2
Outline

• Assumes everyone is aware of what XML is?
• Ontology task background information
• Ontology details
• What is an ontology?
• The ontology development process
• Why Space Surveillance?
• Our ontology approach using XML Schema
• Space Surveillance ontology contents
• Conclusions

3
Ontology Task Background Information

• U.S. Air Force responsible for strategic defense
  C2 systems
• 1.5 Billion contract awarded in Sept. 00
• MITRE briefed USAF on need to architect systems
  that satisfy current C2 requirements, but also
  scale to meet emerging requirements
• Capturing domain semantics vital to
  interoperability
• Task
• Develop preliminary space surveillance ontology
• Provide insight into applicability of XML in
  space domain
• Develop process needed to describe space
  information objects using XML
• Use process as pathfinder for cross-community
  standardization of military space data
• Investigate and leverage related efforts

4
What is an ontology?
Must also understand what the data means.
For interoperability, understanding data syntax
is not enough.
Ontology the definitions of the information
elements, the individual data items, and the
associated inter-relationships
5
Ontology Development Process

• Developed preliminary ontology approach
• Researched relevant XML efforts
• Selected space surveillance as target domain
• Selected XML Schema rather than Document Type
  Definition (DTD) vocabulary
• Participated in XML training
• Developed XML ontology approach using only XML
  Schemas
• Researched XML tools
• Developed space surveillance schema
• Developed ontology demonstration approach
• Researched registration process of
  government-wide XML Registry
• Documented Space Surveillance Ontology

6
Why Space Surveillance?

• Key mission for United States Space Command
  (USSPACECOM)
• Prerequisite for all other operations that use
  space assets
• A current catalog of where things are in space is
  a foundation for the military space domain
• Data content and usage well understood
• Effort not complicated by trying to define how
  data to be used
• Data content, although vital, does not have
  stringent delivery time requirements
• XML tags add overhead
• Didnt want value of XML based approach obscured
  by stringent performance requirements
• Same mission area as related XML demonstration
  effort

7
Space Surveillance Ontology Approach

• Decided to use newer XML schema vocabulary over
  DTDs
• Document Type Definitions (DTDs) in wide use
  today but being supplanted by emerging World Wide
  Web Consortium (W3C) standard called XML Schema
• XML Schema supports strong data typing,
  inheritance, and cross-namespace linking
• Invented new approach for developing the space
  surveillance ontology using only XML Schema
• Approach could be implemented across broad
  spectrum of domains

8
Space Surveillance Ontology XML Schema Approach
9
Space Surveillance Ontology Contents
Data Transformations
XML Schema
Data Presentations
Data structure Meta-data Definitions
Database
Instance of data content
Data View 1
.html
PredefinedQuery (e.g., XSQL)
.xsd
.xml
Data View 2
XSL Processor
.txt
XML Validator
Data Meta-data
Meta- data
Definitions
.html
.xml
.html
10
Space Surveillance Ontology Contents

• Architecture
• SpaceSurveillancesData container element holds 5
  major elements
• Each major element in separate file
• Contains relevant element and attribute
  declarations
• Contains relevant type definitions
• SpaceSurveillanceGlobals file holds type
  definitions and attribute declarations that apply
  across all major elements
• To facilitate reuse
• To isolate items that may change as schema
  evolves
• Metrics
• 93 elements
• 39 user defined types
• 34 attributes (1 attribute group used 16 times)

Schema developed in only 2-3 weeks
11
Space Surveillance Schema Major Element
Relationships
Satellite and
Elset data

are used to generate tasking
Satellite
Tasking
Element Set
Elsets and tasking info
Observations are

are sent to sensors
used to update
Elsets
Sensor
Observation
Sensors collect observations
12
Ontology DemoSensor Tasking Schema
13
Ontology DemoSensor Tasking Element Descriptions
14
Lessons Learned During Development

• An ontology can capture needed data semantics
• The most difficult part of XML schema development
  is determining the major information objects
• XML is not about terseness
• Our ontology approach is pushing the technology
  envelope
• XML Schemas are superior to DTDs
• Schema development is relatively easy with strong
  domain knowledge available
• Reuse is powerful and easy to do with XML schemas
• XML Schema definition language is flexible
• Incremental integration eased development
• Use of optional elements can be powerful
• It is relatively easy to transform XML tagged data

15
Conclusions

• XML is a technology of choice for
  interoperability - but it is only one aspect of
  the total solution
• Internet standard for data structure and data
  exchange
• Enjoys tremendous commercial support (e.g., Sun,
  MS,)
• Provides data transparency
• Does not define semantics or address data storage
• Capturing domain semantics is vital to
  interoperability
• MITRE analyzed ability of XML to capture
  semantics and developed results applicable across
  many domains
• XML Schema vocabulary preferred approach (over
  DTDs)
• Namespaces enable of standard data definitions
• Expect near-term USAF guidance to C2 programs
  requiring definition of information objects using
  XML Schema and registration of tags in XML
  Registry