Alejandro de Frutos Sanchez

1

Using Semantic Web Technologies (OWL) for MDA
Meta-modelling.
Alejandro de Frutos Sanchez
2
Outline

• Introduction
• MDA
• Ontology
• Mapping between Models
• EDOC Ontology
• Benefits
• Current and future work
• Questions

3
Introduction I

• Currently there is a movement towards a new
  approach in the applications development MDA.
• At the same time the Web is giving a new step in
  the content provision, the semantic Web, based on
  Ontologies.

4
Introduction II

• Our work aim to integrate both approaches and get
  the benefits they offer separately.
• The specific application consists in the
  creation of an ontology for describing a modeling
  language and for describing the mapping between
  two models.

5
Model Driven Architecture

• MDA is an open, vendor-neutral approach of the
  OMG group.
• MDA points to separate and relate business or
  application logic from the underlying specific
  platform technology.

6
MDA Structure

• MDA the core of the application development
  process.
• Creation of Platform Independent Model (PIM).
• Middleware platforms.
• Platform Specific Model (PSM).
• There is a need to specify modeling language in
  which the models are defined. MOF.
• Mapping between the Models.

7
Ontology

• Definition of Ontology
• Is a formal explicit description of concepts in
  a domain of discourse. Deborah L. McGuiness
• An ontology typically consists of
• Hierarchical description of important concepts in
  domain.
• Descriptions of properties of each concept.
• An axiomatization that constrains the
  interpretation of these concepts.
• As a domain of discourse we consider the modeling
  languages.

8
The two approaches

• MOF vs OWL
• MOF is the meta-modeling language propose by MDA.
• Our approach is to use OWL to define modeling
  languages.
• Try to infer which benefits and drawbacks it
  involves.

9
Mapping between Models
Generic Mapping Ontology
EDOC Ontology

• Web Services Ontology

Web Services Instance
EDOC Instance
Corba Ontology
Corba Instance
10
EDOC Architecture. CCA
11
CCA Ontology

• Using Owl (Ontology Web Language)
• ltowlClass rdfID"FlowPort"gt
• ltrdfslabel xmllang"en"gtFlowPortlt/rdfslabelgt
  
• ltrdfssubClassOfgt ltowlClass
  rdfabout"Port"/gt
• lt/rdfssubClassOfgt
• lt/owlClassgt
• ltowlFunctionalProperty rdfID"type"gt
• ltrdftyperdfresourceowlObjectProperty"/gt
• ltrdfsdomaingt
• ltowlClassgt
• ltowlunionOf rdfparseType"Collection"gt
• ltowlClass rdfabout"PropertyDefinitio
  n"/gt
• ltowlClass rdfabout"FlowPort"/gt
• lt/owlunionOfgt
• lt/owlClassgt
• lt/rdfsdomaingt
• ltrdfsrange rdfresource"DataElement"/gt
• lt/owlFunctionalPropertygt

12
Benefits using ontologies

• Expressiveness.
• Expressiveness for describing the elements and
  restrictions is better than MOF.
• Make explicit the semantic restrictions.
• And the instances must comply with the ontology
  restrictions.
• Consistency checking.
• Check the consistence of instances of models,
  validation.

13
Current and future work

• Extending the CCA ontology to include the
  composition and choreography parts.
• Working with the reasoner for the validation of
  models. Find out which other features we can
  infer.
• For instance, use the ontology as meta-model
  checker.
• Make a Web Services ontology
• or reuse an existing one.
• Make the mapping ontology between the EDOC and
  Web services Ontologies.
• Or use a general ontology mapping.

14
Questions