People‎ > ‎

Marija Bjeković


Tentative thesis title

Use of modelling languages in enterprise modelling

Problem definition

In enterprise modelling, a wide range of models is typically used to support different purposes. These different models, that jointly represent different perspectives of one enterprise, are also often expressed using different modelling languages. We use the term enterprise modelling landscape to refer to this variety of models and languages used in a specific enterprise modelling effort. The plethora of models/languages used brings about the risk of having fragmented perspectives on the enterprise. This is in turn likely to have a negative impact on the traceability, the ability to do cross-cutting analysis, to manage inconsistency and to ensure overall coherence of the enterprise modelling landscape.

Different approaches are developed to avoid or deal with such a situation, i.e. to achieve model integration. Traditional approach consists in defining an integrated and unifying modelling language, such as e.g. ArchiMate and UML. Moreover, different approaches exist to achieve point-to-point bridges between the different languages. These approaches are based on a priori standardised language definitions, i.e. their syntactic and to some extent semantic aspects. However, in the actual use, one can observe how the ‘standardising’ and ‘integrating’ effect of the language definition as well as these traditional integration approaches erodes. This typically manifests itself by the emergence of dialect-like variations of the original standard language, ‘light weight versions’, or several extensions of an existing standard intended to deal with the ‘missing aspects’.

Instead of designing yet another mechanism to control this phenomena, we argue that it should first be fundamentally understood. To account for the functioning of a modelling language in a socio-pragmatic context of modelling, we claim it is necessary to go beyond a normative view often adopted in the study of modelling languages. 

Research questions

  1.  What are the drivers of defining and standardising fixed modelling languages to be used in enterprise modelling?
  2. Why do the variants of fixed modelling languages emerge in their use in enterprise modelling and how can these variants be characterised? 
  3. How can the linguistic variety be effectively controlled in enterprise modelling?   

Research objective

Develop an explanatory theory as to why and how modelling languages are used in enterprise modelling

Research approach

In this research, we adopt the paradigm of socio-pragmatic constructivism. We understand models as essentially means of representation of some socially constructed knowledge. The modelling process is understood as the process of constructing, representing and sharing this knowledge between the involved participants. Our study of modelling languages goes beyond isolated analysis of the linguistic code, putting forward its function. 

The explanatory theory relies on a conceptual framework on modelling developed as a critical synthesis of the existing theoretical work. 

Research results
  • Bjeković, M. Proper, H.A., Sottet, J.-S.,Towards a coherent enterprise modelling landscape, PoEM (Short Papers), Vol.933, 2012.
  • Bjeković, M. Sottet, J.-S. Favre, J.-M., Proper, H.A., A Framework for Natural Enterprise Modelling, Proceedings of 15th IEEE Conference on Business Informatics (CBI) 2013, p.79-84, 2013.
  • Zarwin, Z. Bjeković, M. Sottet, J.-S., Favre, J.-M., Proper, H.A., Natural Modelling, To appear in Journal of Object Technology, 2014. 

  • Bjeković, M. Proper, H.A. and Sottet, J.-S., Enterprise Modelling Languages - Just Enough Standardisation?, Business Modeling and Software Design, Third International Symposium, BMSD 2013, Revised Selected Papers, LNBIP 173, 2014.

  • Bjeković, M. Proper, H.A. and Sottet, J.-S., Embracing Pragmatics, To be published in E. Yu et al. (Eds.): ER 2014, LNCS 8824, pp. 431--444. Springer, 2014.