Computer Science Colloquium

In this talk we describe the ACT-R cognitive architecture and its application to a cognitive model of a synthetic air traffic control task. ACT-R is a hybrid architecture that combines a symbolic production system with a connectionist-like layer of subsymbolic parameters that continuously adapt to the structure of the environment using Bayesian learning mechanisms. We then describe an ACT-R model of a human operator in a synthetic air traffic control task developed as part of the AMBR modeling competition. The model arises in a principled manner from the structure of the task and the constraints of the architecture. It fits a wide range of quantitative measures of human behavior, including response times, type and quantity of errors, choice probabilities, and subjective workload reports, at the aggregate and individual level. The model generalizes to never previously encountered scenarios and is in many ways undistinguishable from human performance. To conclude, we examine the relation between Artificial Intelligence and Cognitive Modeling in a historical perspective, review a set of criteria proposed by Newell for intelligent architectures and discuss how ACT-R succeeds and fails at meeting these criteria.

Christian Lebiere is a Research Scientist in the Human-Computer Interaction Institute at Carnegie Mellon University. He received his B.S. in Computer Science from the University of Liege (Belgium) and his M.S. and Ph.D. from the School of Computer Science at Carnegie Mellon University. During his graduate career, he worked on the development of connectionist models, including the Cascade-Correlation neural network learning algorithm that has been used in hundreds of scientific, technical and commercial applications. Since 1990, he has worked on the development of the ACT-R hybrid cognitive architecture and is co-author with John R. Anderson of the 1998 book The Atomic Components of Thought. His main research interest is cognitive architectures and their applications to psychology, artificial intelligence, human-computer interaction, decision-making, game theory, and computer-generated forces.