On Spatial Ontologies GEOINFO 2004

1
On Spatial OntologiesGEOINFO 2004

• Stefano Spaccapietra
• Swiss Federal Institute of Technology at Lausanne
  (EPFL)
• http//lbd.epfl.ch

2
You may interrupt me anytime
3
Outline

• Motivation
• On ontologies
• On GIS databases
• On spatial ontologies
• On logic-based approaches
• On conceptual modeling based approaches

4
Changing Focus

• 20th Century Data Processing
• 21st Century Data Exchange
• The fundamental issue has become
  Mutual Understanding
• --gtgt Explicit Semantics
• --gtgt Ontologies (not XML)

5
An Ontology is ...

• "An agreed description of a conceptualization"
• somewhere
  (could be centralized or distributed)
• some set of
  (definitely not necessarily a partition)
• somehow related terms
  (ontology language definition)
• whose use has to some extent been agreed upon
• preferably with some explanation of their meaning
• Ontologies are a means to support semantic
  interoperability
• Ontologies are objects of interest (Universe of
  Discourse), e.g. for ontology management software

6
Without ontologies ...

• How do I know how to interpret
• Where do you come from ? (domain ambiguity)
• Geneva (the airport I started from) ?
• Lausanne or Switzerland (the place Im living in)
  ?
• France (the country I am a citizen of) ?
• Milano (the place I was born) ?
• Ill have a cup of coffee (context dependent)
• Would you consider paying 10000 US to buy a bad
  painting ? (term ambiguity)

7
Ontological Agreement

• Simple case common, shared ontology
• Needs services to define, store, retrieve,
  update, the ontology

8
Cooperative Systems

• Autonomous ontologies

Mediation Ontology
Ontology B
Ontology A
Mediator
A
B
information exchange
9
Taxonomic Ontologies

• sophisticated dictionary/thesaurus
• organized collection of terms
• some semantic links (synonymy, etc.)
• generalization/specialization hierarchy
• example Wordnet
• They provide a reference vocabulary

10
Wordnet
11
Descriptive Ontologies

• concepts are worth a description
• beyond how to denote them (terms)
• which characteristic properties?
• which characteristic relationships?
• They provide information to align existing data
  structures and patterns to define new specialized
  ontologies
• --gt same as conceptual modeling?

12
GIS vs classical databases

• Classification is this water extent a lake or
  lagoon or pond or ?
• Is a PhD student a Student or a Staff member ?
• Is an invited professor a Faculty or not ?
• Fuzzy Boundaries were does the forest stop ?
• When does a ftus start to be a Person ?
• When does a car turn into a wreck ?

13
GIS vs classical DB (ctd.)

• Contextual boundaries depends on whether you see
  a lake or a marsh
• Salary amount depends on whether you see a person
  as an Employee or as a Tax-Payer
• Is a building a feature or an object ?
• Are leather seats a feature of your car or an
  object ?

I have a hard time finding something in
Geo-DB that has no counterpart in non-Geo_DB
14
Spatial Ontologies - 1

• Ontologies of Space
• defining what a point, a line, a point set, etc.
  is
• Open Geospatial Consortium ISO
• Ontologies of Time
• defining what an instant, an interval, a
  duration, etc. is
• ISO Temporal SQL
• Ontologies of Space and Time
• defining what a moving point, a moving line, etc.
  is
• research community

15
Spatial Ontologies - 2

• Ontologies of geographical domains
• water management
• electricity network
• roads, traffic, and transportation
• ........
• These are alike other ontologies
• no specific requirement

16
Spatial Ontologies - 3

• Spatio-Temporal Ontologies (alike
  spatio-temporal DB)
• ontological concepts localized in space and in
  time
• e.g., "soccer", localized in the USAis
  equivalent to "football" localized in
  Europe
• "fat lady" is a kind of Chinese pottery from the
  Han period

17
Supporting Theories

• Logic-based Approaches
• Description Logics
• F(rame) Logics
• Horn Logics
• Conceptual Modeling Approaches

18
Description Logics

• Designed for ontological reasoning
• Many variants (different compromises between
  expressive power, decidability, and complexity of
  reasoning)
• Very popular with the AI/ontology community
• Focuses on axiomatic description of concepts and
  roles (T-box), but also allow description of
  instances (A-box)

19
DL examples

• Primitive concepts
• Concept Person
• Defined concepts
• Committee 10 isCommitteeMember-1.Person
• Author ? Person ? ? writes.Paper
• GoodConference ? Conference ? ? chairedBy.Iochpe

20
F-Logics

• Rule language, designed for deduction
• Object-oriented expressiveness
• person name gt personName,
• firstNames gtgt personFirstname,
• address_at_(type) gt personAddress,
• isMemberOf gtgt committee,
• chairs gtgt committee
• Rule Pchairman PpersonchairsgtgtC

21
Horn-Logics

• Rule languages designed for deduction
• Support recursive rules
• Mostly relational-based (e.g., Datalog)
• reviewer(P,C) pcMember(P,C)
• reviewer(P,C) delegates(Px,P), reviewer(Px,C)

22
Assessment

• DLs
• open world assumption
• automatic consistency checking
• automatic placement of new concepts
• good for distributed asynchronous coordination
• counter-intuitive
• poor expressiveness
• poor readability
• poor query languages
• poor scalability

23
Assessment (2nd)

• F-Logics and Horn-Logics
• close world assumption
• no need for consistency checking
• no automatic placement of new concepts
• can serve as an implementation platform for DL
• (cf. DIP project)

24
Practical perspective

• OWL Ontology Web Language
• RACER a description logic reasoning system
  which implements the SHIQ Logic
• KAON an ontology and semantic web framework
  allowing the design and management of ontologies
• DOGMA an ontology engineering framework based on
  the ORM (Object-Role-Modeling) conceptual model
• Interface Tools Protégé, OntoEdit, ..........

25
Conceptual Modeling Approaches

• DOGMA
• binary model (à la NIAM)
• top ontology domain ontology
• constraints as part of the domain ontologies
• MADS
• spatio-temporal (EER) data model
• AI Reasoning determined Modeling
• DB Modeling first, Reasoning second

26
Logics and Space

• Theoretical ApproachExtend DLs with spatial
  (and temporal) concrete domains
• e.g., polygons (Haarlsev)
• Pragmatic ApproachCombine DLs with GIS (Wessel)

27
Extending DLs

• concrete domains of polygons
• all objects have an associated polygon
• hierarchy of topological relationship
• define concept restrictions
• SwissLake ? Lake ? ? hasArea.g_inside(Switzerland)
  
• define topological roles
• isSInside ? ? (hasArea)(hasArea).strictly_inside
• topological reasoning (but no values)

28
Combining DLs and GIS

• Goal scalability
• Extensional component
• holds instances
• instances are split into thematic part (RACER)
  and geometry part (ad-hoc GIS)
• Intensional component
• DL reasoning
• Query component
• hybrid language

29
Our proposal

• Human-Oriented Approach to Ontology Modeling
• Conceptual Modeling
• MADS spatio-temporal conceptual data model
• Hybrid System
• DBMS DL
• Need for enhancing the reasoning capabilities of
  conceptual models
• symmetric and transitive relationships
• derived objects
• membership predicates
• ......

30
Rationale

• Descriptive ontologies require
• Rich models to enable building representations as
  close as possible of human perception
• Support for the precise definition of concepts in
  relation to other concepts
• Storage and transactions management mechanisms
  (security, concurrency, reliability) to
  realistically manage large sets of instances
• Both open world and closed world reasoning
• Query languages for schema exploration, reasoning
  on the schema, and instance querying

31
Conclusion

• Generic spatial data exchange can become a
  reality as part of the Semantic Web
• Web Services require access to ontologies to
  become Semantic Web Services
• Needed Spatial Ontologies include
• Ontologies of space and time
• Ontologies of geographical domains
• Spatio-Temporal Ontologies
• Human-oriented ontologies need enhanced
  conceptual models to fulfill spatial ontology
  requirements
• It is a long way to go

32
Stefano. Spaccapietra_at_epfl.ch Swiss Federal
Institute of Technology Lausanne (EPFL)