Designing and Enacting Cross-organisational Business Processes: A Model-driven, Ontology-based Approach

Stephan Roser

1st Examiner: Prof. Dr. Bernhard Bauer; 2nd Examiners: Prof. Dr. Jörg P. Müller, Prof. Dr. Wolfgang Reif
Dissertation, Department of Computer Science, University of Augsburg, 2008

Under the pressure of globalization, companies are urged to constantly adapt to new market situations and competitors innovations. Focusing on their core business and core competencies, they engage in Cross-organisational Business Processes (CBPs) with new partners all over the world in ever changing constellations. Companies are organized into global networks and outsource those activities that can be performed more quickly and effectively or at lower costs, by others. These developments create new challenges for enterprise Information and Communication Technology (ICT), requiring ICT systems to support constantly changing enterprise collaboration relationships and to create application systems that support or automate business process enactment starting from business level descriptions and models of CBPs.

Model Driven Software Development (MDSD) provides techniques to realize and automate the propagation of changes at the business level to the technical level. MDSD can be used to provide end-to-end support for the realization of business processes, from the business level (users' view) down to deployed applications (ICT view) on specific platforms via well-defined, largely automated model transformations and refinements. However, there still exist several problems that prevent MDSD from being practically applicable for efficient and effective CBP enactment.

This thesis provides contributions that enable MDSD projects to improve their possible impact on software development and the way ICT systems support business. It develops solutions to three main problems, namely for the areas of CBP modelling and enactment infrastructure, ICT architecture selection, and model and transformation evolution. The contributions of this thesis are as follows:

  1. We develop architecture patterns, a code generation framework, and model transformations that facilitate the generation of executable code from high-level, domain-specific models. These artifacts help to bridge the semantic gap between domain-specific, high-level business process descriptions and the technologies used to implement them in service-oriented ICT systems. Domain and IT experts benefit since our contribution provides a set of sensible and customizable transformations they can reuse to improve the development of their ICT systems.
  2. We investigate new evaluation and decision methods as well as guidelines for the selection of appropriate ICT architectures. We develop a model for decision support that helps IT architects to derive an appropriate architecture paradigm for a given use case or application domain. The decision model combines the Analytic Hierarchy Process (AHP) with scenario-based architecture evaluation techniques. Further, we describe how contingencies influence ICT system coordination architecture. Our decision method, the scenario descriptions, and a set of guidelines help IT architects to select and reuse appropriate ICT system coordination architectures for CBP enactment in a timely manner. This helps to gain productivity wins by reducing the development time and improving the quality of development with existing and tested solutions.
  3. We develop the Ontology-based Model Transformation (OntMT) approach that facilitates the exchange of models between different enterprises as well as the reuse and evolution of model transformations. OntMT helps to overcome differences in syntax and semantics of modelling formalisms with as little effort as possible. Therefore, we apply semantic and reasoning technology to the domain of MDSD. We develop a higher-order model transformation language that allows to modify model transformations. Organisations and modellers benefit from OntMT since it allows them to exchange, customize, and evolve models and model transformations more efficiently.

URN: urn:nbn:de:bvb:384-opus-8057