Data Modeling vs' Ontology Engineering Authored by: Peter Spyns, Robert Meersman, Mustafa Jarrar

1
Data Modeling vs. Ontology EngineeringAuthored
by Peter Spyns, Robert Meersman, Mustafa Jarrar

• Presented by Suprabha Ray
• Date February 2, 2005
• Time 200 pm 450 pm

2
Presentation Outline

• Data models and ontologies similarities and
  differences
• Ontology engineering approach - DOGMA

3
Data Models

• Support quality checks at high levels of
  abstractions
• Provide conceptual constructs like integrity,
  taxonomy, derivation rules
• Information stored diagrammatically Developed for
  in-house usage

4
Ontologies

• Natural language
• Required to share and exchange data as well as
  data semantics in Internet and open connectivity
  environments
• Must represent agreed and shared domain semantics
  for meaningful communication by data exchange and
  interaction of transactions independent of the
  internal technologies of communicating computer
  systems
• Typically used and accessed at runtime

5
Data models vs. ontologiesBoth consist of
conceptual relations and rules

• Data Models
• Offline storage
• Capture semantics for a given application domain

• Ontologies
• Accessible at run-time
• Must contain logical theories for an application
  domain
• Must be reusable and shareable

6
Data models vs. ontologies
continued

• Reference points
• Operation levels
• Data models at a lower implementation oriented
  level
• Ontologies provide constructs for integrity,
  totality
• Expressive powers
• User, purpose and goal relatedness
• Data models capture developers perception of
  domain
• Extendibility

7
Ontology base, Commitments and Lexons the DOGMA
approach

• Ontological resources can be decomposed into
• ontology bases in the form of simple binary facts
  called lexons (4 set tuple - ltg ,T1, R, T2gt
• For each context g and term T, the pair (g, T)
  unique concept.
• ontological commitments in the form of
  description rules and constraints.

8
Ontology base, Commitments and Lexons


continuedExamples of lexons for the
Scientific Conferences Domain
9
Ontology base, Commitments and Lexons

continued

• Example of a commitment by an application that
  will register submitted papers
• (ConferenceAdmin Commitment)
• ltEach Committee ChairedBy at most one Persongt
• ltEach Person who chairs a Committee must also
  IsMemberOf that Committeegt
• ltEach Reviewer Reviews at least one Papergt
• ltEach Paper which is WrittenBy a Person must not
  ReviewedBy with that Persongt

10
Ontology base, Commitments and Lexons

continued
ORM-ML

• Representing ORM models in XML-based syntax so
  files can be accessed and processed at run-time
• Building style-sheets
• Easier schema integration and transformation
• Conceptual queries over the web

11
BiblioOntology Base - Lexons
12
BiblioOntology Base

continued

• Bookstore application
• ISBN identifies a book

• Library application
• Not so no ISBN for journals, manual. Book
  identified by title,author
• Price not relevant to library system

13
Conclusion

• Compare and contrast data models and ontologies ?
  gap between genericity of ontologies and
  specificity of domain rules ? proposal for use of
  DOGMA in ontological engineering ? an
  illustration in support of DOGMA

14
References

• Jarrar M., Demey J., Meersman R. On Using
  Conceptual Data Modeling for Ontology Engineering
  
• Jarrar M. Meersman R., (2002), Formal Ontology
  Engineering in the DOGMA Approach
• Guarino N. Giaretta P., (1995), Ontologies and
  Knowledge Bases Towards a Terminological
  Clarification
• Guarino N., (1998), Formal Ontologies and
  Information Systems

15
Thank you!