An integrated approach to the development of advice systems to support learning and domain based information retrieval

We present a method for developing advice systems. The focus is on allowing users to access information about large, ill-defined domains using their existing knowledge and understanding of those domains. The method relies on iterative refinement and integration of models of the domain, the requirements and the interface. We propose the use of a Generic Advice Systems Architecture which provides the high-level structure for interfaces to such systems, without premature commitment to a particular 'look and feel'. The method is illustrated with material from an ongoing project aimed at developing an advice system to support the integration into Higher Education of students with disabilities.