ABSTRACT

This book has reported the results of our efforts to make ACT–R a theory capable of modeling a wide range of phenomena. The key step in that development has been reducing the size of ACT–R’s knowledge units to what have been advertised as the atomic components of thought As a result of this commitment to a consistent, atomic grain size in knowledge representation, modeling in ACT–R has become much more principled and there has been a substantial convergence on the values for the subsymbolic parameters that control the system. This book described some of the in-house applications of ACT–R as demonstrations of the productivity of this approach. However, these application are but a few of the many that researchers have developed. As a better representation of the range of applications of ACT–R, Table 12.1 lists the research papers presented at the 4th Annual ACT–R Workshop in August 1997. The range of applications is truly gratifying in that it indicates that ACT–R is becoming a useful tool for modeling many aspects of human cognition. This breadth of application is a sign that the ACT–R theory is capturing significant generalizations about the nature of human cognition. Papers Presented at the 1997 ACT–R Summer Workshop

Saturday, August 2

Session 1

Bruno Emond

Models of natural language comprehension and parsing

Eric Scott

Implementing a schema theory in ACT–R

Mike Matessa

Focused learning in linguistic role assignment

Session 2

Kevin Gluck

Learning to learn from a computer-based tutor: An ACT–R model as proof-of-concept

Chris Schunn

Psychologist in a box: An ACT–R model that designs and interprets experiments

Brian Ehret

ACT–R models of submariner situation assessment

Wayne Gray & Erik Altmann

Dynamic microstrategies as an explanation of cognitive workload

Sunday, August 3

Session 1

Tony Simon

Computational evidence for the nonnumerical basis of number competence

Todd Johnson

Computation and retrieval in alphabet arithmetic

Christian Lebiere

Lessons from cognitive arithmetic

Session 2

Marsha Lovett, Lynne Reder, & Christian Lebiere

Modeling working memory effects at the individual level

Dieter Wailach

Modeling complex problem solving

John Anderson & Jonathan Betz

Modeling categorization in ACT–R

Niels Taatgen

Explicit learning in ACT–R

Monday, August 4

Session 1

Joyce Tang Boyland

Modeling syntactic priming in ACT–R

Todd Johnson

Multistrategy learning and transfer in tic-tac-toe

Ken Koedinger & Ben MacLaren

Modeling strategy learning in early algebra

Session 2

Frank Lee

Eye tracking in the air traffic controller task

Dario Salvucci

Relating ACT–R models and eye movement protocols

Tony Simon

Modeling a functional limit to the subitizing phenomenon

Mike Byrne

ACT–R and PRP

Tuesday, August 5

Session 1

Raluca Vasilescu

An ACT–R model for learning anaphoric metaphors

Peter Brusilovsky

ACT–R on the Web

John Anderson, Dan Bothell, Scott Douglass, & Christian Lebiere

ACT–R models of the navigation task