Specialisation and Generalisation of Processes

  • Christine Choppy LIPN, CNRS UMR 7030, Université Paris 13, Université Sorbonne Paris Cité, France
  • Jörg Desel FernUniversität in Hagen, Germany
  • Laure Petrucci LIPN, CNRS UMR 7030, Université Paris 13, Université Sorbonne Paris Cité, France
Keywords: Process Modelling, Process Generalisation, Process Specialisation

Abstract

In conceptual data modelling, one of the most important abstraction concepts is specialisation, with generalisation being the converse. Although there are already some approaches to define generalisation for behavioural modelling as well, there is no generally accepted notion of process generalisation. In this paper, we introduce a general definition of process specialisation and generalisation. Instead of concentrating on a specific process description language, we refer to labelled partial orders. For most process description languages, behaviour (if defined at all) can be expressed by means of this formalism. We distinguish generalisation from aggregation, and specialisation from instantiation. For Petri nets, we
provide examples and suggest associated notations. Our generalisation notion captures various previous approaches to generalisation, for example ignoring tasks, allowing alternative tasks and deferring choices between alternative tasks. A general guideline is that a more general process contains less features and/or less information than a more specific one.

Published
2018-02-27
Section
Invited Contribution