Formal Ontology and Information Systems

1
Formal Ontology and Information Systems

• Nicola Guarino (FOIS98)
• Presenter Yihong Ding
• CS652 Spring 2004

2
Ontologies are important

• Knowledge engineering
• Knowledge representation
• Language engineering
• Qualitative modeling
• Information modeling
• Information retrieval, extraction, and
  integration
• Object-oriented analysis
• Knowledge management and organization
• Database design
• Agent-based system design
• The Semantic Web

3
e-Business requests

• Trying to engage with too many partners too
  fast is one of the main reasons that so many
  online market makers have foundered. The
  transactions they had viewed as simple and
  routine actually involved many subtle
  distinctions in terminology and meaning
• Harvard Business Review, October 2001

4
Technical problem for e-Business requests

• Lack of technologies and products to
  dynamically mediate discrepancies in business
  semantics will limit the adoption of advanced Web
  services for large public communities whose
  participants have disparate business processes
• Gartner Research, February 28, 2002

5
XML is not the solution

• XML is only the first step to ensuring that
  computers can communicate freely. XML is an
  alphabet for computers and as everyone who
  travels in Europe knows, knowing the alphabet
  doesnt mean you can speak Italian or French
• Business Week, March 18, 2002

6
Open- and closed-world assumptions

• Closed-world assumption
• The information provided is complete (a knowledge
  base contains all relevant facts).
• Known the knowledge base is incomplete (does not
  have enough information to produce an answer to a
  question), a decision must be made without
  complete information.
• If you cannot prove P or not P, assume it is
  false.
• This is the usual semantics of relational
  databases.
• The closed-world assumption is designed to
  finesse but not solve these problems and is
  adopted in default of a better solution.

7
Open- and closed-world assumptions

• Open-world assumption
• Any proposition or theorem which cannot be
  derived from the facts and axioms present in the
  system is held to be unknown.
• Things which are known to be true or false must
  be stated explicitly, or else inferable from
  facts and axioms.
• The two boolean values (true and false) are
  inadequate, and we have to use the
  ThreeValuedLogic.
• The open-world assumption more clearly models
  reality.
• The number of domains can be infinite.

8
Example from schema to ontology
Local Database
9
Example from schema to ontology
10
What is an ontology?

• An ontology is a formal, explicit specification
  of a shared conceptualization. Gruber 93
• Formal
• The ontology should be machine readable.
• Explicit
• The type of concepts use, and the constraints of
  their use are explicitly defined.
• Shared
• The ontology should capture consensual knowledge
  accepted by the communities.
• Conceptualization
• An ontology is an abstract model of phenomena in
  the world by having identified the relevant
  concepts of those phenomena.

11
What is a conceptualization?

• Formal structure of (a piece of) reality as
  perceived and organized by an agent,
  independently of
• the vocabulary used
• the actual occurrence of a specific situation

12
Conceptualization
Conceptualization of scene 1 lta, b, c, d, e ,
on, above, clear, table gt
13
Relations vs. Conceptual Relations
ordinary (extensional) relations are defined on a
domain D conceptual (intensional) relations are
defined on a domain space ltD, Wgt

• A conceptualization is a set of conceptual
  relations defined on a domain space.

14
Intended Model and Ontological Commitment

• World structure a structure of ltD, Rgt, which
  refers to a particular world
• Intended world structure a world structure for a
  conceptualization in a particular world
• Each conceptualization contains many of them.
• One intended world structure for each world.
• Intended model the representation of an intended
  world structure in a model by ontological
  commitment.
• Ontological commitment the intensional
  interpretation of a logical language L

15
Ontologies and Intended Models
Ontology
16
Ontology Quality
imprecision
incompleteness
17
The Ontology Integration/Sharing Problem (1)
Agents A and B can communicate only if their
intended models overlap
18
The Ontology Integration/Sharing Problem (2)
Two different ontologies may overlap while their
intended models do not (especially if the
ontologies are not accurate enough)
19
The role of foundational ontologies (1)
IB(L)
False agreement minimized
IA(L)
20
The role of foundational ontologies (2)

• Bottom-up integration of domain-specific
  ontologies can never guarantee consistency of
  intended models (despite apparent logical
  consistency).
• Top-level foundational ontologies
• Simplify domain-specific ontology design
• Increase quality and understandability
• Encourage reuse

21
Hierarchies of Ontologies
22
Towards Ontology-Driven IS temporal dimension
(1)

• Using an ontology at development time
• Benefit
• Enable knowledge reuse instead of software reuse
• Enable application domain knowledge reuse and
  share across heterogeneous software platforms
• Avoid bothering too much on implementation
  details
• Two scenarios
• First scenario has ontology library containing
  reusable domain and task ontologies
• Second scenario has very generic ontology
  consisting of coarse domain-level distinctions

23
Towards Ontology-Driven IS temporal dimension
(2)

• Using an ontology at run time
• Benefit
• Enable communication between software agents
• Two cases
• Ontology-aware IS an IS component is just aware
  of the existence of a (possibly remote) ontology
  and can query it for whatever specific
  application purpose
• Ontology-driven IS the ontology is just another
  component (typically local to the IS),
  cooperating at run time towards the higher
  overall IS goal

24
Towards Ontology-Driven IS structural dimension

• Ontology as a database component
• Ontology as a user interface component
• Ontology as an application program component