Skip to main content Accessibility help
×
  • Cited by 43
Publisher:
Cambridge University Press
Online publication date:
June 2012
Print publication year:
2008
Online ISBN:
9780511816772

Book description

This book is a definitive reference source for the growing, increasingly more important, and interdisciplinary field of computational cognitive modeling, that is, computational psychology. It combines breadth of coverage with definitive statements by leading scientists in this field. Research in computational cognitive modeling explores the essence of cognition and various cognitive functionalities through developing detailed, process-based understanding by specifying computational mechanisms, structures, and processes. Given the complexity of the human mind and its manifestation in behavioral flexibility, process-based computational models may be necessary to explicate and elucidate the intricate details of the mind. The key to understanding cognitive processes is often in fine details. Computational models provide algorithmic specificity: detailed, exactly specified, and carefully thought-out steps, arranged in precise yet flexible sequences. These models provide both conceptual clarity and precision at the same time. This book substantiates this approach through overviews and many examples.

Reviews

"[...] This edited volume by Sun (Rensselaer Polytechnic Institute) comprises two sections:[...] The first section will attract broader interest, especially from students, because of its juxtaposition of distinct approaches including connectionist, Bayesian, and logical modeling. The second section covers a range of topics, from memory and learning to decision making and cognitive control. [...] Given that the application chapters are largely independent of the methodological chapters, a dedicated instructor could cover more extensive ground by selecting primary papers on a desired topic. However, researchers who use computational approaches, or who want to become better consumers of computational psychology literature, may find this to be a valuable compilation of major ideas in this area. Recommended.
--S.A. Huettel, Duke University CHOICE

"--With the publication of The Cambridge Handbook of Computational Psychology, the newly emerging, interdisciplinary field of computational cognitive modeling has come of age...a cutting-edge overview of classic and currentwork in computational psychology. This handbook stakes out this important and promising area of cognitive science...a definitive reference source for therapidly growing, increasingly important, and strongly interdisciplinary field ofcomputational cognitive modeling...The Cambridge Handbook of Computational Psychology represents a milestone, marking a number of important contributions to the larger field of cognitive science."
--Howard T. Everson, PsycCRITIQUES [May 20, 2009, Vol. 54, Release 20, Article 5]

Refine List

Actions for selected content:

Select all | Deselect all
  • View selected items
  • Export citations
  • Download PDF (zip)
  • Save to Kindle
  • Save to Dropbox
  • Save to Google Drive

Save Search

You can save your searches here and later view and run them again in "My saved searches".

Please provide a title, maximum of 40 characters.
×

Contents


Page 1 of 2


  • 5 - Declarative/Logic-Based Cognitive Modeling
    pp 127-169
  • View abstract

    Summary

    This chapter provides an overview of the book on Introduction to Computational Cognitive Modeling. The first part of the book provides a general introduction to the field of computational cognitive modeling. The second part, Cognitive Modeling Paradigms, introduces the reader to broadly influential approaches in cognitive modeling. The third part, Computational Modeling of Various Cognitive Functionalities and Domains, describes a range of computational modeling efforts that researchers in this field have undertaken regarding major cognitive functionalities and domains. This part surveys and explains computational modeling research, in terms of detailed computational mechanisms and processes, on memory, concepts, learning, reasoning, decision making, skills, vision, motor control, language, development, scientific explanation, social interaction, and so on. The final part, Concluding Remarks, explores a range of issues associated with computational cognitive modeling and cognitive architectures, and provides some perspectives, evaluations, and assessments.
  • Part III - Computational Modeling of Various Cognitive Functionalities and Domains
    pp 187-664
  • View abstract

    Summary

    In this chapter, computer models of cognition focusing on the use of neural networks are reviewed. This chapter begins by placing connectionism in its historical context, leading up to its formalization in Rumelhart and Mc-Clelland's two-volume Parallel Distributed Processing. Three important early models illustrating some of the key properties of connectionist systems are discussed, as well as how the novel theoretical contributions of these models arose from their key computational properties. Connectionism offers an explanation of human cognition because instances of behavior in particular cognitive domains can be explained with respect to a set of general principles and the conditions of the specific domains. Connectionist theory has had a widespread influence on cognitive theorizing, and this influence was illustrated by considering connectionist contributions to our understanding of memory, cognitive development, acquired cognitive impairments, and developmental deficit.
  • 7 - Computational Models of Episodic Memory
    pp 189-225
  • View abstract

    Summary

    The chapter focuses on problems in higher-level cognition: inferring causal structure from patterns of statistical correlation, learning about categories and hidden properties of objects, and learning the meanings of words. This chapter discusses the basic principles that underlie Bayesian models of cognition and several advanced techniques for probabilistic modeling and inference coming out of recent work in computer science and statistics. The first step is to summarize the logic of Bayesian inference based on probabilistic models. A discussion is then provided of three recent innovations that make it easier to define and use probabilistic models of complex domains: graphical models, hierarchical Bayesian models, and Markov chain Monte Carlo. The central ideas behind each of these techniques is illustrated by considering a detailed cognitive modeling application, drawn from causal learning, property induction, and language modeling, respectively.
  • 8 - Computational Models of Semantic Memory
    pp 226-266
  • View abstract

    Summary

    The dynamical systems approach to cognition is the theoretical framework within which this embodied view of cognition can be formalized. This chapter reviews the core concepts of the dynamical systems approach and illustrates them through a set of experimentally accessible examples. Particular attention is given to how cognition can be understood in terms that are compatible with principles of neural function, most prominently, with the space-time continuity of neural processes. The chapter reviews efforts to form concepts based on the mathematical theory of dynamical systems into a rigorous scientific approach toward cognition that embraces the embodied and situated stance. The chapter explains how behavioral signatures of the neural field dynamics may provide evidence for the Dynamical Field Theory (DFT) account of cognition. The theoretical concept of stability, at the core of dynamical systems thinking, is the key to understanding autonomy.
  • 9 - Models of Categorization
    pp 267-301
  • View abstract

    Summary

    This chapter talks about systematization of a particular approach to modeling the mind: declarative computational cognitive modeling. The goal of computational cognitive modeling and the goal of declarative computational cognitive modeling and systematization in logic-based computational cognitive modeling (LCCM) are to understand the kind of cognition distinctive of human persons by modeling this cognition in information processing systems. LCCM is made based on a generalized form of the concept of logical system as defined rather narrowly in mathematical logic. This chapter shows how the problems can be solved in LCCM in a manner that matches the human normatively incorrect and normatively correct responses returned after the relevant stimuli are presented. This chapter explains LCCM as a formal rationalization of declarative computational cognitive modeling. It also presents the attempt to build computational simulations of all, or large portions of, human cognition, on the basis of logic alone.
  • 10 - Micro-Process Models of Decision Making
    pp 302-321
  • View abstract

    Summary

    Cognitive architectures are on the one hand echoes of the original goal of creating an intelligent machine faithful to human intelligence and on the other hand attempts at theoretical unification in the field of cognitive psychology. This chapter discusses the current state of cognitive architectures to characterize four prime examples: The States, Operators, And Reasoning (SOAR) architecture, the Adaptive Control of Thought, Rational (ACT-R) theory, Executive-Process Interactive Control (EPIC) architecture, and Connectionist Learning with Adaptive Rule Induction Online (CLARION) architecture. The chapter examines a number of topics that can serve as constraints on modeling and discusses how four architectures offer solutions to help modeling in that topic area. The viewpoint of cognitive constraint is different from the perspective of how much functionality an architecture can provide, as expressed by, for example, Anderson and Lebiere.
  • 12 - Mental Logic, Mental Models, and Simulations of Human Deductive Reasoning
    pp 339-358
  • View abstract

    Summary

    The term episodic memory refers to the ability to recall previously experienced events and to recognize things as having been encountered previously. Research on the neural basis of episodic memory has increasingly come to focus on three structures: The hippocampus, Perirhinal cortex and Prefrontal cortex. This chapter reviews the Complementary Learning Systems (CLS) model and how it has been applied to understanding hippocampal and neocortical contributions to episodic memory. In addition to the biologically based models, there is a rich tradition of researchers building more abstract computational models of episodic memory. The chapter describes an abstract modeling framework, the Temporal Context Model (TCM) that has proved to be very useful in understanding how to selectively retrieve memories from a particular temporal context in free recall experiments. Episodic memory modeling has a long tradition of trying to build comprehensive models that can simultaneously account for multiple recall and recognition findings.
  • 13 - Computational Models of Skill Acquisition
    pp 359-395
  • View abstract

    Summary

    Understanding the basis of the human abilities, to recognize, comprehend, and make inferences about objects and events in the world and to comprehend and produce statements about them, is the goal of research in semantic memory. Semantic memory is memory for meanings. In some disciplines (e.g., linguistics), the word semantics refers exclusively to the meanings of words and sentences. Collins and Quillian's model effectively uses the syllogism as a basis for organizing propositional knowledge in memory. The chapter offers a promising theoretical framework for semantic cognition. The chapter provides a simple framework for thinking about how coherent covariation between linguistic structure and other aspects of experience can promote the representation of meaning for full sentences and events. The chapter explores patterns of semantic impairment to reveal the neuroanatomical organization of the semantic system.
  • 14 - Computational Models of Implicit Learning
    pp 396-421
  • View abstract

    Summary

    This chapter surveys a variety of formal models of categorization, with emphasis on exemplar models. The chapter reviews exemplar models' similarity functions, learning algorithms, mechanisms for exemplar recruitment, formalizations of response probability, and response dynamics. A mutual goal of different formal models is to account for detailed quantitative data from laboratory experiments in categorization. Although a variety of representational formats have been formalized, exemplar models have been especially richly explored by researchers. The chapter discriminates numerous exemplar models to excise their functional components and to examine those components side by side. The main functional components include the computation of similarity, the learning of associations and attention, the recruitment of exemplars, the determination of response probability, and the generation of response times. This dissection revealed a variety of formalizations available for expressing any given psychological process.
  • 15 - Computational Models of Attention and Cognitive Control
    pp 422-450
  • View abstract

    Summary

    Computational models are like the new kids in town for the field of decision making. This field is dominated by axiomatic utility theories or simple heuristic rule models. Decision theory has a long history, starting as early as the seventeenth century with probabilistic theories of gambling by Blaise Pascal and Pierre Fermat. In an attempt to retain the basic utility framework, constraints on utility theories are being relaxed, and the formulas are becoming more deformed. Recently, many researchers have responded to the growing corpus of phenomena that challenge traditional utility models by applying wholly different approaches. This chapter provides concrete illustration of how the computational approach can account for all of the behavioral paradoxes that have contested utility theories. The extent to which the other computational models have been successful in accounting for the results is also discussed.
  • 16 - Computational Models of Developmental Psychology
    pp 451-476
  • View abstract

    Summary

    This chapter addresses one important aspect of inductive reasoning, namely, psychological research on category-based induction, or how people use categories to make likely inferences. It describes similarity effects, typicality effects, diversity effects, and other phenomena, including background knowledge effects, setting the stage for the presentation of computational models of inductive reasoning. One consideration to keep in mind as computational models are presented is whether they have any facility for addressing not only similarity, typicality, and diversity effects, but also background knowledge effects and indeed whether they show any capacity for causal reasoning. The chapter discusses two general issues that arise in modeling inductive reasoning and also in computational modeling of other cognitive activities. The first issue is that cognitive activities do not fall neatly into pigeonholes. The second is that putting background knowledge into models is the necessary next step.
  • 17 - Computational Models of Psycholinguistics
    pp 477-504
  • View abstract

    Summary

    Individuals, who know no logic, are able to make deductive inferences. For many years, psychologists argued that deduction depends on an unconscious system of formal rules of inference akin to those in proof-theoretic logic. The first mental model theory is for simple inferences based on quantifiers, and programs have simulated various versions of this theory, and the probabilistic theory often makes unsatisfactory predictions. The theory of mental models posits that the engine of human reasoning relies on content. The simulation of model theory concerns sentential reasoning, and it shows how an apparently unexceptional assumption leads to a striking prediction of systematic fallacies in reasoning - a case that yields crucial predictions about the nature of human deductive reasoning. The chapter concludes with an attempt to weigh up the nature of human rationality in the light of other simulation programs.
  • 18 - Computational Models in Personality and Social Psychology
    pp 505-529
  • View abstract

    Summary

    This chapter focuses on cognitive as opposed to sensori-motor skills and on models that create or alter symbolic knowledge representations. It deals briefly with models that learn by adjusting quantitative properties of knowledge structures. Although occasionally referring to empirical studies, the chapter is primarily a review of theoretical concepts. It proceeds on the assumption that each hypothesis contains some grain of truth to be extracted and incorporated into future models. The learning mechanism is a more finegrained unit than the model or the cognitive architecture. Cognitive descriptions of processes in the mind are functional descriptions of what this or that piece of wetware is doing, what function it carries out. This perspective points to the need to understand the relation between learning mechanisms and modes of neural plasticity.
  • 19 - Cognitive Social Simulation
    pp 530-548
  • View abstract

    Summary

    Learning is implicit when an individual acquires new information without intending to do so. The distinction between implicit and explicit knowledge may hinge on whether a person is conscious of the regularity with a conscious rather than unconscious mental state. Computational modeling has played a central role in deconstructing early verbal theories of the nature of what is learned in implicit learning paradigms. On the theoretical and conceptual applications of implicit learning, this chapter addresses three central issues: whether performance in implicit learning situations result in abstract knowledge; whether the data and the modeling suggest the involvement of single or multiple systems; and whether modeling is relevant to addressing the conscious versus unconscious nature of the acquired knowledge. Implicit learning has proven to be a rich domain for exploration of the differences between information processing with and without consciousness.
  • 20 - Models of Scientific Explanation
    pp 549-564
  • View abstract

    Summary

    The study of attention is central to understanding how information is processed in cognitive systems. Modern cognitive research interprets attention as the capacity to select and enhance limited aspects of currently processed information. This chapter reviews key computational models and theoretical directions pursued by researchers trying to understand the multifaceted phenomenon of attention. A broad division is drawn between theories and models addressing the mechanisms by which attention modulates specific aspects of perception (primarily visual) and those that have focused on goal-driven and task-oriented components of attention. An area of recent activity in elaborating on the computational mechanisms of goal-driven attention concerns mechanisms by which attentional biases arise or are modulated during the course of task performance. Finally, the chapter focuses on the contrast or continuum between attentional control and automaticity, an issue that becomes crystallized when examining the distinctions between, or transitions from, novice to expert cognitive task performance.
  • 21 - Cognitive Modeling for Cognitive Engineering
    pp 565-588
  • View abstract

    Summary

    This chapter provides a comparative survey of computational models of psychological development. To understand how computational modeling can contribute to the study of psychological development, it is important to appreciate the enduring issues in developmental psychology. The most common computational techniques applied to psychological development are production systems, connectionist networks, dynamic systems, robotics, and Bayesian inference. The chapter discusses modeling in the areas of the balance scale, past tense, object permanence, artificial syntax, similarity-to-correlation shifts in category learning, discrimination-shift learning, concept and word learning, and abnormal development. Some of the models reviewed in this chapter simulated development with programmer designed parameter changes. Variations in such parameter settings were used to implement age-related changes in both connectionist and dynamic-systems models of the A-not-B error, the Cascade-Correlation (CC) model of discrimination-shift learning, all three models of the similarity-to-correlation shift, and the autism model.

Page 1 of 2


Metrics

Altmetric attention score

Full text views

Total number of HTML views: 0
Total number of PDF views: 0 *
Loading metrics...

Book summary page views

Total views: 0 *
Loading metrics...

* Views captured on Cambridge Core between #date#. This data will be updated every 24 hours.

Usage data cannot currently be displayed.