Skip to main content

Computing by Graph Transformation


The major aim of COMPUGRAPH II is to demonstrate the potential of graph transformation as a uniform framework for the development of modern software systems. In order to reach this aim, objectives have been established under the headings of foundations, concurrent computing, and graph transformation for specification and programming.

Within the area of foundations, research is directed towards the unification, extension, and completion of existing graph transformation theories.

In concurrent computing, the aim is to elaborate a graph transformation-based approach which can provide operation semantics for various models of concurrent computing (including Petri-nets and process description languages) and enhances correct and comprehensive specifications.

The graph transformation for specification and programming area considers graph transformation as the kernel of a rule-based specification language and aims at extending this kernel by typing mechanism, module and transformation concepts, applicability constraints, transactions, and concepts developed within concurrent computing in order to improve the feasibility of graph rewriting as a system specification method.
The research is working towards a unification and extension of the theory of graph transformation so that it can provide a uniform paradigm for computing, that is, for the specification, implementation, programming, and analysis of sequential, parallel and distributed systems.

A Hybrid Query Language for extended entity relationship databases has been developed. The language integrates graphical and textual notations. Its semantics is based on graph transformations. There is an ongoing research of applying high level replacement systems to different kinds of petri-nets. This has already lead to important new results for net transformation systems with interesting applications to specification of distributed systems.
A suitable variant of graph grammars has been used to specify the operational behaviour of an abstract machine that implements narrowing.
The notion of an actor grammar has been generalized to the more natural notion of an electronic support measure (ESM) system for which a compositional semantics has been presented. A different graph transformation model of actor systems has motivated the investigation of transformations dealing with consistent graph structures only.
A new framework for attributed graph transformations has been proposed. The manipulation of values taken from a suitably specified algebra is smoothly integrated into the framework of single pushout graph transformations.
Investigations about a new, truly concurrent semantics for graph grammars has been started. The first relevant result is the definition of an original notion of equivalence among graph derivations, which relates derivations which are essentially the same from the truly concurrent point of view.
Collage grammars have been introduced as a rule based device for picture generation. Collages consist of geometric objects together with entities 'hyperedges' that are repeatedly replaced by collages during the generation process.

The work within COMPUGRAPH II is structured around three research areas: foundations; concurrent computing; graph transformation for specification and programming. Smaller teams have been established to tackle area-based issues.

This work will be organised by joint research and several meetings of two or more partners in each team in cooperation with European and international experts; for this reason the Group has drawn up an non-exclusive list of 11 scientific correspondents.

Three international workshops - including one in the USA in 1994 (for which an additional joint ESPRIT/NSF proposal has just been accepted) - will be used to demonstrate and integrate the results of the research areas. Moreover results will also be presented in scientific journals and at European and international conferences. COMPUGRAPH II will participate in and contribute to workshops of ESPRIT groups investigating related issues, including SEMAGRAPH II (6345), COMPASS II (6112), and workshops and conferences in the area of theoretical and applied Computer Science.


COMPUGRAPH II should provide a solid semantic foundation for new system specification and programming paradigms based on unified graph transformation models and the necessary theoretical background for initiating languageand CASE-based projects centered around these paradigms.


Salzufer 17-19
10587 Berlin

Participants (5)

Rijnsburgerweg, 10, 9600
2300 RC Leiden
Bibliothekstrasse 1
28334 Bremen
Università degli Studi di Pisa
Lungarno Pacinotti 45
56100 Pisa
Université de Bordeaux I
351 Cours De La Libération
33405 Talence
Pleinlaan, 2
1050 Bruxelles