China
Africa
Explore chapters and articles related to this topic
Theory and Practice From Cognitive Science
Published in Constantine Stephanidis, User Interfaces for All, 2000
The chunking hypothesis and the associated view of learning have been used as one of the main components in the design of the SOAR problem-solving architecture. SOAR is a problem space theory of human cognitive architecture proposed by Laird et al. (1987) as an integrative theory. In contrast to the specific experimental studies and limited theories described so far, SOAR was used by Howes and Young (1996) to construct an integrated model of learning and performance at the user interface (TAL).
Model-computer interaction: Implementing the action perception loop for cognitive models
Published in Don Harris, Engineering Psychology and Cognitive Ergonomics, 2017
Gordon D. Baxter, Frank E. Ritter
Two models which incorporate interaction with their application have been built to date. These models were implemented using the Soar cognitive architecture. The simulations of the tasks, and the perceptual mechanisms, were developed using Garnet (Myers, Giuse, Dannenberg, Vander Zanden, Kosbie, Pervin, Mickish, & Marchal, 1990). The first application was a simulation of a simplified Air Traffic Control (ATC) task. The second simulation was a simple tabletop on which a number of blocks were located.
Architecture
Published in David Burden, Maggi Savin-Baden, Virtual Humans, 2019
David Burden, Maggi Savin-Baden
SOAR (https://soar.eecs.umich.edu/), developed in the early 80s by John Laird, Allen Newell, and Paul Rosenbloom at Carnegie Mellon University, was an early attempt to develop a practical cognitive architecture which could be implemented as software to provide behaviours that mirrored those of human cognition.
World models and predictive coding for cognitive and developmental robotics: frontiers and challenges
Published in Advanced Robotics, 2023
Tadahiro Taniguchi, Shingo Murata, Masahiro Suzuki, Dimitri Ognibene, Pablo Lanillos, Emre Ugur, Lorenzo Jamone, Tomoaki Nakamura, Alejandra Ciria, Bruno Lara, Giovanni Pezzulo
In cognitive science, cognitive functionalities such as memory, perception, and decision-making are implemented as modules in the cognitive architectures studied, and the specific task can be solved by activating these modules coordinately. ACT-R [251] and Soar [252] are representatives of cognitive architectures. It has been shown that the model implemented by ACT-R can explain the time to solve the task by humans, and activation patterns of the brain can be predicted by activation patterns of the modules [253]. Furthermore, Soar has been used for controlling robots [254] and learning games [255]. However, complex machine learning methods that have rapidly advanced in a decade are not introduced yet. Sigma [256, 257] is a newer cognitive architecture that introduces the generative flow graph, a generalized probabilistic graphical model. Therefore, the model can be implemented using probabilistic programming techniques [258–260]. Furthermore, the concept of the standard model of the mind is discussed through a synthesis across these three cognitive architectures [237]. Particularly, cognitive architectures based on first principles, e.g. with a general computation scheme, such as free energy minimization [19], are especially attractive. The architecture for social cognition has also been proposed [261]. The authors point out that these architectures explained above are incomplete in dealing with the social aspect of cognition and describe the elements of architecture for social cognition. Clarion [262] is another cognitive architecture based on dual process theory [263]. In this architecture, each subsystem is composed of explicit and implicit processes, and it is shown that the interaction between implicit-explicit processes can explain psychological phenomena.