Documents
Resources
Learning Center
Upload
Plans & pricing Sign in
Sign Out

Ontology-Driven Spatial Information Retrieval in GRIP - CISC

VIEWS: 4 PAGES: 52

									   Ontology-driven spatial
information retrieval in GRIP



 Naicong Li, Nathan Strout, Steve Paplanus, Thomas Leuteritz
                    University of Redlands
                      Redlands, CA USA

                  AAG 2007, San Francisco
Ontology-driven spatial information retrieval in GRIP


  •   Objective
  •   Geographically Referenced Information Portal (GRIP)
  •   Desert tortoise recovery plan data repository
  •   Information retrieval requirement analysis
  •   System design
  •   Future work
Ontology-driven spatial information retrieval in GRIP


   Objective
        Knowledge organization and information retrieval for data with
         heterogeneous format, syntax, and semantics
       – Use GRIP as development platform, use desert tortoise recovery plan
         data repository as project data
  •   Geographically Referenced Information Portal (GRIP)
  •   Desert tortoise recovery plan data repository
  •   Information retrieval requirement analysis
  •   System design
  •   Future work
Ontology-driven spatial information retrieval in GRIP


   Objective
       – Knowledge organization and information retrieval for data with
         heterogeneous format, syntax, and semantics
        Use GRIP as development platform, use desert tortoise recovery plan
         data repository for project data
  •   Geographically Referenced Information Portal (GRIP)
  •   Desert tortoise recovery plan data repository
  •   Information retrieval requirement analysis
  •   System design
  •   Future work
Ontology-driven spatial information retrieval in GRIP


  • Objective
   Geographically Referenced Information Portal (GRIP)
        Open service architecture
       – Supports distributed processing
       – Search data with heterogeneous data format
  •   Desert tortoise recovery plan data repository
  •   Information retrieval requirement analysis
  •   System design
  •   Future work
Ontology-driven spatial information retrieval in GRIP


  • Objective
   Geographically Referenced Information Portal (GRIP)
       – Open service architecture
        Supports distributed processing
       – Search data with heterogeneous data format
  •   Desert tortoise recovery plan data repository
  •   Information retrieval requirement analysis
  •   System design
  •   Future work
Ontology-driven spatial information retrieval in GRIP


  • Objective
   Geographically Referenced Information Portal (GRIP)
       – Open service architecture
       – Supports distributed processing
        Search data with heterogeneous data format
  •   Desert tortoise recovery plan data repository
  •   Information retrieval requirement analysis
  •   System design
  •   Future work
Ontology-driven spatial information retrieval in GRIP


  • Objective
  • Geographically Referenced Information Portal (GRIP)
   Desert tortoise recovery plan data repository
        Desert tortoise declared endangered species
       – USFWS recovery plan identifying critical habitat units (CHU) and recommended
         recover actions
       – CHUs under management of various agencies: BLM, NPS, military bases, etc.
       – Centralized data repository with digital library and GIS data
       – GRIP to provide framework underlying the information portal for the data
         repository
  •   Information retrieval requirement analysis
  •   System design
  •   Future work
Ontology-driven spatial information retrieval in GRIP


  • Objective
  • Geographically Referenced Information Portal (GRIP)
   Desert tortoise recovery plan data repository
       – Desert tortoise declared endangered species
        USFWS recovery plan identifying critical habitat units (CHU) and recommended
         recover actions
       – CHUs under management of various agencies: BLM, NPS, military bases, etc.
       – Centralized data repository with digital library and GIS data
       – GRIP to provide framework underlying the information portal for the data
         repository
  •   Information retrieval requirement analysis
  •   System design
  •   Future work
Ontology-driven spatial information retrieval in GRIP


  • Objective
  • Geographically Referenced Information Portal (GRIP)
   Desert tortoise recovery plan data repository
       – Desert tortoise declared endangered species
       – USFWS recovery plan identifying critical habitat units (CHU) and recommended
         recover actions
        CHUs under management of various agencies: BLM, NPS, military bases, etc.
       – Centralized data repository with digital library and GIS data
       – GRIP to provide framework underlying the information portal for the data
         repository
  •   Information retrieval requirement analysis
  •   System design
  •   Future work
Ontology-driven spatial information retrieval in GRIP


  • Objective
  • Geographically Referenced Information Portal (GRIP)
   Desert tortoise recovery plan data repository
       – Desert tortoise declared endangered species
       – USFWS recovery plan identifying critical habitat units (CHU) and recommended
         recover actions
       – CHUs under management of various agencies: BLM, NPS, military bases, etc.
        Centralized data repository with digital library and GIS data
       – GRIP to provide framework underlying the information portal for the data
         repository
  •   Information retrieval requirement analysis
  •   System design
  •   Future work
Ontology-driven spatial information retrieval in GRIP


  • Objective
  • Geographically Referenced Information Portal (GRIP)
   Desert tortoise recovery plan data repository
       – Desert tortoise declared endangered species
       – USFWS recovery plan identifying critical habitat units (CHU) and recommended
         recover actions
       – CHUs under management of various agencies: BLM, NPS, military bases, etc.
       – Centralized data repository with digital library and GIS data
        GRIP to provide framework underlying the information portal for the data
         repository
  •   Information retrieval requirement analysis
  •   System design
  •   Future work
    GIS DATA – SDE GEOSPATIAL DATABASE


•   147 tortoise related feature classes
•   501 base map feature classes
•   26 raster data sets
•   All with FGDC compliant metadata




                         •ESRI             •ESRI
                   DTP RESOURCE LIBRARY

•   The DTP Library is a searchable repository in Sharepoint, combining all
    the project resources. The digital library includes 4000 items.
•   The documents are sorted in the following categories:
     •   Free form text documents
     •   Images
     •   Maps
Ontology-driven spatial information retrieval in GRIP


  •   Objective
  •   Geographically Referenced Information Portal (GRIP)
  •   Desert tortoise recovery plan data repository
     Information retrieval requirement analysis
          Improve interoperability among heterogeneous data
       –   Allow integrated queries for GIS and non GIS data
       –   Organize query results semantically
       –   Facilitate knowledge discovery
  •   System design
  •   Future work
Ontology-driven spatial information retrieval in GRIP


  •   Objective
  •   Geographically Referenced Information Portal (GRIP)
  •   Desert tortoise recovery plan data repository
     Information retrieval requirement analysis
       –   Improve interoperability among heterogeneous data
          Allow integrated queries for GIS and non GIS data
       –   Organize query results semantically
       –   Facilitate knowledge discovery
  •   System design
  •   Future work
Ontology-driven spatial information retrieval in GRIP


  •   Objective
  •   Geographically Referenced Information Portal (GRIP)
  •   Desert tortoise recovery plan data repository
     Information retrieval requirement analysis
       –   Improve interoperability among heterogeneous data
       –   Allow integrated queries for GIS and non GIS data
          Organize query results semantically
       –   Facilitate knowledge discovery
  •   System design
  •   Future work
Ontology-driven spatial information retrieval in GRIP


  •   Objective
  •   Geographically Referenced Information Portal (GRIP)
  •   Desert tortoise recovery plan data repository
     Information retrieval requirement analysis
       –   Improve interoperability among heterogeneous data
       –   Allow integrated queries for GIS and non GIS data
       –   Organize query results semantically
          Facilitate knowledge discovery
  •   System design
  •   Future work
Ontology-driven spatial information retrieval in GRIP


  •   Objective
  •   Geographically Referenced Information Portal (GRIP)
  •   Desert tortoise recovery plan data repository
  •   Information retrieval requirement analysis
     System design
          Integrated solution to knowledge organization and information retrieval
       –   Ontology development
       –   Ontology-based information organization
       –   Ontology-based information retrieval
       –   Making implicit information in data explicit
  •   Future work
Ontology-driven spatial information retrieval in GRIP

   System design -- Integrated solution to knowledge organization and
    information retrieval
Ontology-driven spatial information retrieval in GRIP

  •   System design -- Integrated solution to knowledge organization and
      information retrieval

                                               Knowledge organization
Ontology-driven spatial information retrieval in GRIP

    •   System design -- Integrated solution to knowledge organization and
        information retrieval




 Information retrieval
Ontology-driven spatial information retrieval in GRIP


   System design -- Ontology development

       Application domain specific and project specific
Ontology-driven spatial information retrieval in GRIP


   System design -- Ontology development

      • Application domain specific and project specific
       Based on a controlled vocabulary agreed upon by a user
        community
          –   Prototype ontology being developed for system design and testing purposes
Ontology-driven spatial information retrieval in GRIP


   System design -- Ontology development

      •   Application domain specific and project specific
      •   Based on a set of controlled vocabulary agreed upon by a user
          community
           –   Prototype ontology being developed for system design and testing purposes
       Categories, instances, attributes, relations
Ontology-driven spatial information retrieval in GRIP


   System design -- Ontology development

      •   Application domain specific and project specific
      •   Based on a set of controlled vocabulary agreed upon by a user
          community
           –   Prototype ontology being developed for system design and testing purposes
      • Categories, instances, attributes, relations
       All the concepts have corresponding natural language
        expression(s)
Ontology-driven spatial information retrieval in GRIP


   System design -- Ontology development

      •   Application domain specific and project specific
      •   Based on a set of controlled vocabulary agreed upon by a user
          community
           –   Prototype ontology being developed for system design and testing purposes
      • Categories, instances, attributes, relations
      • All the concepts have corresponding natural language
        expression(s)
       Axioms defined based on project needs
Ontology-driven spatial information retrieval in GRIP


   System design -- Ontology development

      •   Application domain specific and project specific
      •   Based on a set of controlled vocabulary agreed upon by a user
          community
           –   Prototype ontology being developed for system design and testing purposes
      • Categories, instances, attributes, relations
      • All the concepts have corresponding natural language
        expression(s)
      • Axioms defined based on project need
       In OWL-DL
Ontology-driven spatial information retrieval in GRIP


   System design -- Ontology development

      •   Application domain specific and project specific
      •   Based on a set of controlled vocabulary agreed upon by a user
          community
           –   Prototype ontology being developed for system design and testing purposes
      • Categories, instances, attributes, relations
      • All the concepts have corresponding natural language
        expression(s)
      • Axioms defined based on project need
      • In OWL-DL
       Using SWEET as upper level ontology
Ontology-driven spatial information retrieval in GRIP


                              Natural language expressions




                            Attribute and relation definitions
Ontology-driven spatial information retrieval in GRIP



                       Natural language expression




                                   Attribute and relation definitions

                       Instances
Ontology-driven spatial information retrieval in GRIP



                     Causal relationship




                     Threat factor categories
Ontology-driven spatial information retrieval in GRIP


     Example of inverse relations:
       •   manages – managedBy
       •   predatorOf – hasPredator

     Example of transitive relations: causal relation
       • Between threat factors and tortoise population reduction
       • Among threat factors
Ontology-driven spatial information retrieval in GRIP


     Example of inverse relations:
        •       manages – managedBy
        •       predatorOf – hasPredator

     Example of transitive relations: causal relation
       • Between threat factors and tortoise population reduction
       • Among threat factors
                 supports            causes
     Landfill               ravens            tortoise population reduction
Ontology-driven spatial information retrieval in GRIP


     Example of inverse relations:
        •       manages – managedBy
        •       predatorOf – hasPredator

     Example of transitive relations: causal relation
       • Between threat factors and tortoise population reduction
       • Among threat factors
                 supports            causes
     Landfill               ravens            tortoise population reduction
                  causes                          causes
     Landfill               habitat extent loss            tortoise population reduction
Ontology-driven spatial information retrieval in GRIP


     Example of inverse relations:
        •       manages – managedBy
        •       predatorOf – hasPredator

     Example of transitive relations: causal relation
       • Between threat factors and tortoise population reduction
       • Among threat factors
                 supports            causes
     Landfill               ravens            tortoise population reduction
                  causes                          causes
     Landfill               habitat extent loss            tortoise population reduction

                  causes                          causes
     Grazing                habitat extent loss            tortoise population reduction
Ontology-driven spatial information retrieval in GRIP


     Example of inverse relations:
        •       manages – managedBy
        •       predatorOf – hasPredator

     Example of transitive relations: causal relation
       • Between threat factors and tortoise population reduction
       • Among threat factors
                 supports            causes
     Landfill               ravens            tortoise population reduction
                  causes                          causes
     Landfill               habitat extent loss            tortoise population reduction

                  causes                          causes
     Grazing                habitat extent loss            tortoise population reduction

                  causes                      causes
     Grazing             soil compaction           habitat degradation
                  causes
                         tortoise population reduction
Ontology-driven spatial information retrieval in GRIP


  •   Objective
  •   Geographically Referenced Information Portal (GRIP)
  •   Desert tortoise recovery plan data repository
  •   Information retrieval requirement analysis
     System design
       –   Integrated solution to knowledge organization and information retrieval
       –   Ontology development
          Ontology-based information organization
       –   Ontology-based information retrieval
       –   Making implicit information in data explicit
  •   Future work
Ontology-driven spatial information retrieval in GRIP


   System design -- Ontology-based information organization

       Mapping between concepts in the ontologies and
          –   data in geodatabases
          –   Text documents (keyword field in metadata)
Ontology-driven spatial information retrieval in GRIP


   System design -- Ontology-based information organization

      •   Mapping between concepts in the ontologies and
           –   data in geodatabases
           –   Text documents (keyword field in metadata)


       “Registration mappings” = “a set of rules describing the
        association between semantic types and structural types”*
           –   Xquery for xml data source
           –   SQL expression for relational database data source




      * Bowers & Ludascher 2004
Ontology-driven spatial information retrieval in GRIP


   System design -- Ontology-based information organization

      •   Mapping between concepts in the ontologies and
           –   data in geodatabases
           –   Text documents (keyword field in metadata)


      •   “Registration mappings” = “a set of rules describing the
          association between semantic types and structural types”*
           –   Xquery for xml data source
           –   SQL expression for relational database data source


       Including mapping between process/service concepts and
        geoprocessing tools**

      * Bowers & Ludascher 2004
      ** Lutz & Klein 2006
Ontology-driven spatial information retrieval in GRIP


  •   Objective
  •   Geographically Referenced Information Portal (GRIP)
  •   Desert tortoise recovery plan data repository
  •   Information retrieval requirement analysis
     System design
       –   Integrated solution to knowledge organization and information retrieval
       –   Ontology development
       –   Ontology-based information organization
          Ontology-based information retrieval
       –   Making implicit information in data explicit
  •   Future work
Ontology-driven spatial information retrieval in GRIP


   System design -- Ontology-based information retrieval

       User query types
            Browsing (guided by ontologies)*
           – Searching (query keyword  concept in ontologies; query
             expansion guided by ontologies)
      •   Retrieving GIS and non GIS data through the same query UI
      •   Query results organized semantically by the ontology




      * Guarino 1998, Fonseca et al 2002
Ontology-driven spatial information retrieval in GRIP


   System design -- Ontology-based information retrieval

       User query types
           – Browsing (guided by ontologies)
            Searching (query keyword  concept in ontologies; query
             expansion guided by ontologies*)
      •   Retrieving GIS and non GIS data through the same query UI
      •   Query results organized semantically by the ontology




      * Fu et al 2005
Ontology-driven spatial information retrieval in GRIP




        Live demo for design concept
Ontology-driven spatial information retrieval in GRIP


   System design -- Ontology-based information retrieval

      •   User query types
           – Browsing (guided by ontologies)
           – Searching (query expansion guided by ontologies)
       Retrieving GIS and non GIS data through the same query UI
      • Query results organized semantically by the ontology
Ontology-driven spatial information retrieval in GRIP


   System design -- Ontology-based information retrieval

      •   User query types
           – Browsing (guided by ontologies)
           – Searching (query expansion guided by ontologies)
      • Retrieving GIS and non GIS data through the same query UI
       Query results organized semantically by the ontology
Ontology-driven spatial information retrieval in GRIP


   System design -- Ontology-based information retrieval

      •   User query types
           – Browsing (guided by ontologies)
           – Searching (query expansion guided by ontologies)
      •   Retrieving GIS and non GIS data through the same query UI
      •   Query results organized semantically by the ontology

   System design -- Ontology-based information retrieval
      Making implicit information in data explicit
Ontology-driven spatial information retrieval in GRIP


  •   Objective
  •   Geographically Referenced Information Portal (GRIP)
  •   Desert tortoise recovery plan data repository
  •   Information retrieval requirement analysis
     System design
  •   Future work
Ontology-driven spatial information retrieval in GRIP


  •   Objective
  •   Geographically Referenced Information Portal (GRIP)
  •   Desert tortoise recovery plan data repository
  •   Information retrieval requirement analysis
  •   System design
     Future work
Ontology-driven spatial information retrieval in GRIP

  Future Work

  •   Implement the prototype
  •   Registration mapping
  •   Automatic symbolization during GIS layer display
  •   UI for GIS layer ordering, show/hide option
  •   Full text annotation
  •   Document relevance ranking
  •   Applying NLP for text disambiguation
Naicong Li, Nathan Strout, Steve Paplanus, Thomas Leuteritz
                Naicong_Li@institute.redlands.edu
               Nathan_Strout@institute.redlands.edu
              Steve_Paplanus@institute.redlands.edu
             Thomas_Leuteritz@institute.redlands.edu




                     Redlands Institute
                   University of Redlands
                    1200 E. Colton Ave.
                    Redlands, CA 92373

                         909-748-8268

            www.redlands.edu/redlandsinstitute

								
To top