AMODEUS has three aims: (A) To establish a set of related integrational frameworks for expressing different aspects of the design of user-system interaction. (B) To extend the scope of basic modelling techniques to provide analytic leverage on the problems of user-system interaction with sophisticated interactive technologies; and (C) To assess how integrational frameworks, modelling techniques and the substantive principles they convey may successfully be transferred to and used by the design community.
Interdisciplinary approaches are being developed in human computer interaction (HCI) and how such approaches might best be transferred to, and applied by, interface designers is being studied.
The system modelling research has produced a classification of multimodal systems; a comparison of influential architectural models; and has developed an initial theory of interactors. The user modelling research has implemented a prototype programmable user model; developed techniques for analysing user cognition with dynamic graphics together with an analysis of the mental blending of multimodal information; and has developed an initial categorization of user related issues in usability engineering. Integrational and design related research has developed large design spaces and techniques for design space analysis; has clarified key issues for the definition of concepts in interactional semantics; and has developed a preliminary high level categorization of potential design commitments. With respect to the transfer of concepts and techniques to design settings, the project has systematically decomposed aspects of the broader transfer problem; the potential of computer based learning and fuzzy logic techniques have been reviewed; and 2 preliminary empirical tests of transfer possibilities have been carried out.
APPROACH AND METHODS
As a truly interdisciplinary enterprise, the work of the project naturally draws upon a significant range of representational and empirical methods. These include descriptive taxonomies in system, user and design domains as well as an analysis of the gulfs between research tools and design practise; empirical observation and contrastive experiments for studying users and designers; formal methods for representing abstract system models and techniques for representing and contrasting software architectures; production system techniques and cognitive task analysis techniques for modelling user cognition; formal and semi-formal techniques for representing interactions, design space analysis, design processes and domain-based requirements. Transfer techniques under consideration include fuzzy logic and computer-based learning.
Although this represents terrific methodological diversity, by their very nature some of the representational techniques are intended to integrate over other sources. A unique feature of the AMODEUS consortium is the commitment of all groups to applying their methods to the analysis of the same problems in significant sized design spaces. The products of such conjoint analyses form the basis for integrating and transferring basic science into design contexts.
By conducting basic research on representational techniques, their properties and inter-relationships, and their prospects for application in commercial design, the project should lay a firm foundation for the migration of research findings downstream into ESPRIT and other European and national research programmes. This foundation will be communicated through scientific reports and proof of concept demonstrations of the various techniques in the context of concrete examples.
CB2 1AB Cambridge
YO1 5DD York