Cognitive Architectures in HCI: Present Work and Future Directions

This paper serves as the overview and introduction to a symposium of the same name. The symposium is made up of this introduction and six other papers on cognitive architectures in HCI. As many readers may not be familiar with cognitive architectures, a description of what cognitive architectures are is presented first. In an effort to be accessible to a wide audience, this description is fairly abstract. Once it is clear what is meant by a cognitive architecture, and what a model derived from such an architecture is, then the potential uses of such models in HCI efforts, both research and practical, can be laid out. While there is a great deal of promise, there are still challenges involved with using cognitive models in HCI, which are detailed in the third section. Finally, an overview of the other symposium papers is presented along with some orienting context. 1 What are Cognitive Architectures? There are multiple ways to answer this question. In one sense, a cognitive architecture is a theoretical entity. That is, it is a broad theory of human cognition based on a wide selection of experimental data. Cognitive psychology is replete with theories about particular aspects of human cognition; a cognitive architecture is an attempt to build an integrated theory that encompasses a broad spectrum of what is known about human cognition and performance. Cognitive psychology has, for many years, been driven primarily by a divide-and-conquer approach, so cognitive architectures represent something of a departure