Titel: Semantic Cognition: A Parallel Distributed Processing Approach
Autor/en: Timothy T. Rogers, James L. McClelland
Empfohlen ab 18 Jahre.
Januar 2006 - kartoniert - 448 Seiten
This groundbreaking monograph offers a mechanistic theory of the representation and use of semantic knowledge, integrating the strengths and overcoming many of the weaknesses of hierarchical, categorization-based approaches, similarity-based approaches, and the approach often called "theory theory." Building on earlier models by Geoffrey Hinton in the 1980s and David Rumelhart in the early 1990s, the authors propose that performance in semantic tasks arises through the propagation of graded signals in a system of interconnected processing units. The representations used in performing these tasks are patterns of activation across units, governed by weighted connections among them. Semantic knowledge is acquired through the gradual adjustment of the strengths of these connections in the course of day-to-day experience.The authors show how a simple computational model proposed by Rumelhart exhibits a progressive differentiation of conceptual knowledge, paralleling aspects of cognitive development seen in the work of Frank Keil and Jean Mandler. The authors extend the model to address aspects of conceptual knowledge acquisition in infancy, disintegration of conceptual knowledge in dementia, "basic-level" effects and their interaction with expertise, and many findings introduced to support the idea that semantic cognition is guided by naive, domain-specific theories.
Timothy T. Rogers is a research scientist at the Medical Research Council Cognition and Brain Sciences Unit in Cambridge, England. James L. McClelland is Professor of Psychology and Director of the Center for Mind, Brain, and Computation at Stanford University. He is the coauthor of Parallel Distributed Processing (1986) and Semantic Cognition (2004), both published by the MIT Press.
"A fascinating exploration of the conceptual systems that emerge when a neural network is trained to predict the properties of objects. Rogers and McClelland show that incremental, error-driven learning leads to internal distributed representations that explain a whole variety of empirical phenomena. Their network exhibits remarkable common sense in the way it generalizes, and distinctly human frailty in the way its knowledge disintegrates when it is physically degraded. Conceptual tendencies that many researchers assume to be innate are produced by the interplay between error-driven learning and the higher-order statistical structure of the set of facts that the network learns. The book uses very little technical jargon and the reasoning is clear, detailed, and compelling."--Geoffrey Hinton, FRS, Canada Research Chair in Machine Learning, Department of Computer Science, University of TorontoPlease note: The fourth sentence may be omitted for space purposes. "This book -- by one of the founders of the field of computational developmental psychology -- provides a comprehensive and thorough overview of the field, as well as an articulate and persuasive statement of Shultz's own approach. Both beginning students and advanced researchers will find it stimulating, informative, and thought-provoking."--Jeff Elman, Professor of Cognitive Science, University of California, San Diego