Related Communities:

InfoSem Project

INTAS Grant INTAS-94-1817

Information System Semantic Interoperability
(INFOSEM)

Project starting date

July 1995

Project duration

36 months

Project co-ordinator

Prof. E.J. Neuhold
GMD -- German National Research Center for Information Technology,
Integrated Publication and Information Systems Institute,
Darmstadt, Germany

Participants

  • Institute for Problemds of Informatics of the Russian Academy of Science, Moscow, Russia (Leonid Kalinichenko)
  • Kiev State University, Ukraine (Nikolaj Nikitchenko)
  • Versailles University, PRISM Laboratory, France (Georges Gardarin)
  • Integrated Publication and Information Systems Institute, GMD, Darmstadt, Germany (Erick Neuhold)

Objectives

The primary objectives of the INFOSEM project are to investigate, prototype and disseminate methodologies and basic techniques allowing construction of semantically interoperable information systems based on the pre-existing heterogeneous information resources.

A modelling methodology and language facilities meeting the requirements of information resource / application specification completeness, equivalence and uniformity will be developed. Basic framework and discipline leading to semantic interoperation will constitute the methodology for refinement of the application specification by the resource patterns in the interoperable environment. Sound foundations necessary for justification of validity of information system specification concretization by the coherent resource compositions will be incorporated.

The project is specifically oriented towards database-intensive systems. However, semantic interoperation of arbitrary information resources from industrial frameworks will be also supported.

Results obtained

The overall architecture of the sematically interoperable information systems design environment has been defined over the collection of models including ontological model, requirement planning / analysis model, design model, implementation model, information resource specification model. One and the same canonical semi-formal object model (SYNTHESIS) has been defined to be used for different architecture layers.

To give the canonical model exact meaning, a mapping of this object model into the formal Abstract Machine Notation providing precise meaning for the language has been constructed. Thus, we get the semi-formal object model and its formal counterpart that we can use together as a common paradigm for uniform specifications of pre-existing components and for different modeling tasks used on various phases of the information systems development.

Unified Modeling Language (UML) has been chosen as a graphical notation for all phases of the process of development. We supply this notation with a clean object semantics given by the canonical object model.

The architectural framework for component-based information systems development is designed so that the SYNTHESIS method itself could be neutral to existing Object Analysis and Design (OAD) methods. Forward and Backward phases of the methods have been augmented with specific SYNTHESIS facilities, such as ontological specifications needed to resolve contextual differences with the pre-existing resources and complete specifications of types in the canonical object model possessing a possibility to justify reuse on the basis of formal specification and proof facilities.

A method and a process for the semantically interoperable systems design has been developed. We apply a model of composite objects integrating data and behaviour from various sources. The data and /or behaviour residing at each source are regarded as fragments of the composite object. We consider type reducts as the basic patterns of reuse. So, we decompose type specifications into a collection of reducts that can serve as units of reuse and composition. Thus we specify types of the resource fragments as reducts of the required type and provide their composition in a bottom up fashion. Federated composite objects are created by integration of such fragments using views above original resource type definitions.

Type refinement is identified in the project as the fundamental property that is required to justify a correctness of substitution of fragments of one type for another so that we can reuse pre-existing objects state and behaviour. An approach for discovery of suitable reducts of resource types refining reducts of analysis model types to form a design type refining the required type has been developed.

Method for integration of ontologies of the specification of requirements and of an information resource has been developed. In process of the integration, coherence and interrelationship of concepts defined on different levels is established. The integrated ontological specification provides necessary basis for search of the information resource type (class) specifications and their components suitable for the concretization of application type (class).

The SYNTHESIS design procedure has been structured as an autonomous phase. The metamodel to represent SYNTHESIS entities in the metabase supporting this specific phase has been defined. The implementation phase of the method has been specifically designed to be used in the CORBA-based environment. The prototype of the design and implementation phases of the method has been developed.

Related Publications

Supported by Synthesis Group