China
Africa
Explore chapters and articles related to this topic
Architecture
Published in David Burden, Maggi Savin-Baden, Virtual Humans, 2019
David Burden, Maggi Savin-Baden
Hart and Goertzel (2008) divide an AI system design into four components. Cognitive Architecture: the overall design of an AI system; what parts does it have, and how do they connect to each other.Knowledge Representation: how the AI system internally stores declarative, procedural and episodic knowledge, and how creates its own representations for elemental and abstracted knowledge in new domains it encounters.Learning: how the AI system learns new elemental and abstracted knowledge, and how it learns how to learn.Teaching Methodologies: how the AI system is coupled with other systems so as to enable it to gain new knowledge about itself, the world and others.
Model-Based Evaluation
Published in Julie A. Jacko, The Human–Computer Interaction Handbook, 2012
Led by Anderson (1983) and Newell (1990), researchers in human cognition and performance began to construct models using a cognitive architecture (see Byrne, this volume). Cognitive architecture parallels the concept of computer architecture: a cognitive architecture specifies a set of fixed mechanisms, the “hardware,” that comprise the human mind. To construct a model for a specific task, the researcher “programs” the architecture by specifying a psychological strategy for doing the task, the “software” (specifying parameter value settings and information in memory might be involved as well.) The architecture provides the coherent theoretical framework within which the processes and constraints can be proposed and given an explicit and rigorous definition. Several proposed cognitive architectures exist in the form of computer simulation packages in which programming the architecture is done in the form of production systems, collections of modular if-then rules, which have proved to be an especially good theoretical model of human procedural knowledge. Developing these architectures, and demonstrating their utility, is a continuing research activity (see Byrne, this volume). Not surprisingly, they all have a long way to go before they accurately incorporate even a subset of the human abilities and limitations that appear in an HCI design context.
Measuring Up: Benefits and Trends in Performance Measurement Technologies
Published in Christopher Best, George Galanis, James Kerry, Robert Sottilare, Fundamental Issues in Defense Training and Simulation, 2013
Beth F. Wheeler Atkinson, Robert G. Abbott, Danielle C. Merket
Cognitive architectures, such as ACT-R and SOAR (Lehman, Laird, & Rosenbloom, 2006; Anderson, 1996), seek to meet this need by emulating human cognition, such as memory recall, pattern recognition, and planning. They provide custom programming languages tailored for inputting facts and the logical connections between facts. These architectures provide general thinking capabilities, but must be programmed to serve specific roles, such as automated student assessment and instruction. Programming a cognitive architecture requires both technical and domain expertise, and is a large and expensive undertaking. Even then, the system will not have human-level abilities in listening to students and explaining new information to them.
Why is a computational framework for motivational and metacognitive control needed?
Published in Journal of Experimental & Theoretical Artificial Intelligence, 2018
More sophisticated goal structures and (sub)routine mechanisms are needed in cognitive architectures. For instance, in a more sophisticated cognitive architecture, goals may emerge from competitions among different needs and desires, goals may change in various ways (including in a stack-like fashion as well as other possibilities) and so on. Routines may have both of the following two properties: persistence and interruptibility (Tyrell, 1993); the interplay of these properties may need to be addressed (Sun, 2009, 2016). In addition, the initiation of routines (e.g. by setting goals), the routines themselves, and the termination of routines can all be learned (in addition to being hand-coded, as in many existing cognitive architectures). Their learning may be done through autonomous trial-and-error exploration, instructions, imitations and other means (e.g. Sun, 2016; Sun & Sessions, 2000).
AI-enabled Enterprise Information Systems for Manufacturing
Published in Enterprise Information Systems, 2022
Milan Zdravković, Hervé Panetto, Georg Weichhart
A general approach to (human-like) reasoning are cognitive architectures (A. Newell 1990). Cognitive architectures aim at research in general AI – and implement human-like reasoning mechanisms. The goal is to understand how humans think, by providing a formalism (the architecture) which allows mimicking human thinking. Programs written using cognitive architectures (in their final version) would be able to reason across many domains and adapt to new situations and are able to reflect on themselves. Cognitive architecture research supports cognitive science by providing detailed (executable models) of human ‘thinking’ processes (A. Newell 1990).
Cognitive model of phonological awareness focusing on errors and formation process through Shiritori
Published in Advanced Robotics, 2022
Jumpei Nishikawa, Junya Morita
In contrast to such a simple connectionism approach, studies have been conducted using cognitive architecture for cognitive modeling. Cognitive architecture is a design specification that integrates intelligence-generating structures (brain structures) and intellectual functions (e.g. thought processes with regard to specific tasks) [12]. It can be regarded as the foundation (structure) of cognitive modeling, which accumulates cognitive functions used in various tasks. A model using cognitive architecture allows us to isolate factors involved in accomplishing a task using a general structure.