The development of a decision support system generator via action research

The research reported in this thesis has the dual objectives of investigating: 1) the possibility of extending the field of application of visual modelling to decision areas where discrete event simulation is not the appropriate modelling technique; 2) the potential for designing a generalised framework that could form the basis for the development of decision making aids for different problems. The work consisted of the design, development and implementation of a Decision Support System (DSS) for the production planning function of the sponsoring organisation, from which basis it was possible to generalise and isolate the components that form the generator with which other specific applications can be built more effectively. This thesis reflects the action research methodology followed in the project. Some of the chapters are dedicated to the problems associated with the specific application, i.e. the description of the problem and its environment, the issues of managing a multi-disciplinary team with action and research objectives, and the problems of design, implementation and evaluation of a decision making aid in an organisation. The remaining chapters discuss the design principles that were followed in the development of the DSS generator, which is independent of the data base and allows for the possibility of personalised interfaces and a modular structure of models that can grow to match the evolving requirements of different decision makers. From this action research programme I have shown that: 1) It is possible to develop personalized interfaces that can be tuned to the cognitive requirements of different users. 2) The concept of semi-structured decisions can become operational with the proposed new definition of models and searching processes, and this enables a DSS to evolve with the modelling needs of decision makers. This research programme has established design principles associated with those two topics, which should be considered by an analyst who wishes to develop a specific DSS