Abstract
This workshop will focus on appropriate use of cognitive models for the analysis and solution of HCI
problems.
Keywords:
Cognitive modeling, User modeling, Simulation
Introduction
The use of cognitive architectures for modeling is beginning to have an interesting, but little noted, impact
on the types of
modeling applied to HCI issues. In the past, the software used to construct models was either very generic
(as in the use of
OPS5) or handcrafted by the individual modeler. Recently this has changed. At first generic connectionist
architectures and,
more recently, symbol manipulation architectures (Soar and ACT-R) have entered the public domain as
systems that are
centrally supported but whose use has spread beyond their developers. Most of the models presented at
recent CHI
conferences have been developed and presented as an instance of a modeling architecture (e.g., Soar-PUMS
or Soar-GOMS).
The latest generation of models incorporate large amounts of cognitive theory and enable the modeler to
build more
generative and realistic models that take into account more features of human cognition, the artifact
(software design) being
modeled, and the task to which the artifact is being applied. Controlling this C-A-T (cognition-artifact-task)
triad promises
to be the key to more realistic, powerful, and productive models.
The purpose of this workshop is to bring together three overlapping groups of HCI researchers: those who
have used more
than one of the architectures, those who have used just one major architecture, and
those who have
developed cognitive models using either generic or handcrafted software. The aim is to
identify ways in which
the
existence and choice of cognitive architectures bears on the practice and prospects for modeling in HCI.
ACTIVITIES
To provide depth and perspective on the topic, participants will be individuals who (1) have developed a
computer-based,
cognitive model, (2) have a history of research in HCI issues, and (3) are interested in the effect of their
choice of
architecture on the way their models have developed.
Three pre-workshop activities are designed to provide a common frame of reference. First, participants will
provide a
written description of one of the models that they have developed to be shared with other workshop
participants. Second,
three articles (from a list nominated by the participants) were read by all, before coming to the workshop.
Third, participants
will be asked to consider how their approach to modeling would be applied to model one of a small number
of benchmark
tasks.
During the workshop, presentations and exercises will use these models and readings to focus discussions
on appropriate use
of cognitive models for analysis and solution of HCI problems. Specific topics include:
- Which models are best for what kinds of analyses and problems
- Methodology appropriate with the development and use of cognitive models
- Comparison of cognitive modeling and other approaches for solving similar HCI problems
- The theoretic basis for candidate cognitive models
Results of this workshop will serve as a foundation for the scientifically valid and appropriate use of
cognitive modeling and
cognitive architectures for future generations of HCI practitioners. One product of the workshop will be a
report that will be
published in the SIGCHI Bulletin. An expected product is the development of a support-group of
researchers interested in
modeling HCI phenomena.
