<< 1 >>
Rating: 
Summary: FYI, the long Table of Contents
Review: 1 Dimensions 1.1 The Problem of Modeling Representations 1.2 Conceptual Spaces as a Framework for Representations 1.3 Quality Dimensions 1.4 Phenomenal and Scientific Interpretations of Dimensions 1.5 Three Sensory Examples: Color, Sound, and Taste 1.6 Some Mathematical Notions 1.7 How Dimensions Are Identified 1.8 Integral and Separable Dimensions 1.9 On the Origins of Quality Dimensions 1.10 Conclusion 2 Symbolic, Conceptual, and Subconceptual Representations 2.1 An Analogy for the Three Kinds of Representations 2.2 Symbolic Representations 2.3 Subconceptual Representations 2.4 Conceptual Representations 2.5 Connections to Neuroscience 2.6 Comparisons 2.7 The Jungle of Representations 3 Properties 3.1 Program 3.2 Properties in Intensional Semantics 3.3 Criticism of the Traditional View of Properties 3.4 Criteria for Natural Regions of Conceptual Spaces 3.5 Natural Properties 3.6 Reconsidering the Problems 3.7 The Relativism of Conceptual Spaces 3.8 Connections to Prototype Theory 3.9 Voronoi Tessellations of a Space 3.10 Higher Level Properties and Relations 3.11 Conclusion 4 Concepts 4.1 Concepts versus Properties 4.2 Modeling Concepts 4.3 The Role of Similarity in Concept Formation 4.4 Combining Concepts 4.5 Learning Concepts 4.6 Nonmonotonic Aspects of Concepts 4.7 Concept Dynamics and Nonmonotonic Reasoning 4.8 Objects as Special Kinds of Concepts 4.9 Four Geometric Categorization Models 4.10 The Shell Space 4.11 Experiments 5 Semantics 5.1 What Is a Semantics? 5.2 Six Tenets of Cognitive Semantics 5.3 Analysis of Some Aspects of Lexical Semantics 5.4 An Analysis of Metaphors 5.5 The Learnability Question 5.6 Communicating Referents 5.7 Can Meanings Be in the Head? 5.8 Conclusion: The Semantic Program 6 Induction 6.1 Three Levels of Induction 6.2 The Symbolic Level 6.3 The Conceptual Level 6.4 The Role of Theoretical Concepts 6.5 The Subconceptual Level 6.6 Correlations between Domains 6.7 Conclusion: What Is Induction? 7 Computational Aspects 7.1 Computational Strategies on the Three Levels 7.2 Conceptual Spaces as Emergent Systems 7.3 Smolensky's Treatment of Connectionism 7.4 Not All Computation Is Done by Turing Machines 7.5 A System for Object Recognition 7.6 Conclusion 8 In Chase of Space 8.1 What Has Been Achieved? 8.2 Connections among Levels 8.3 Conclusion: The Need for a New Methodology Notes References Illustration Credits Index
Rating:
Summary: An eye opener
Review: For anyone interested in the cognitive topics, machine learning and artificial intelligence, this book is an eye opener. The point of view it presents attempts to put an order in what "meaning" really means. Drawbacks of the book? The lack of conceptualization when it comes to dynamic concepts (treated very superficially). Also, the theory is deficient when modeling the functional aspects of concepts (a "sin" already recognized by the author). But considering the pioneering character of this piece of art, these drawbacks are just compelling invitations for further research in the field.
Rating:
Summary: An eye opener
Review: For anyone interested in the cognitive topics, machine learning and artificial intelligence, this book is an eye opener. The point of view it presents attempts to put an order in what "meaning" really means. Drawbacks of the book? The lack of conceptualization when it comes to dynamic concepts (treated very superficially). Also, the theory is deficient when modeling the functional aspects of concepts (a "sin" already recognized by the author). But considering the pioneering character of this piece of art, these drawbacks are just compelling invitations for further research in the field.
Rating:
Summary: Excellent and Enlightening
Review: Gardenfors introduces his theory of concept-formation, and at the same time presents a survey of the competing theories and research. He shows a high level of professionalism by accepting that the theories can coexist, presenting the competing theories in their strongest light and letting you decide where to apply each theory. This book is not only a good argument for why his theory deserves a permanent place in your toolbox, but also a good education for anyone wanting to know the tradeoffs in representing concepts -- especially for knowledge representation or machine learning systems. He presents the material in a very logical order so that the subtopics can be consumed individually. And although some of the material is well-known, each chapter presents a series of contrasting pros and cons and synthesizes the information in ways that are thought-provoking and novel. It was well worth the time and money.
Rating: 
Summary: A little disappointing
Review: If one is to design a machine that can formulate concepts and engage in such things as inductive inference and its corollary scientific discovery, then one must be able to quantify the notion of a concept in such a way that it can be implemented into the cognitive structure of the machine. One must be able to distinguish one concept from another, be able to tell when one concept is similar to another, and understand in detail how concepts are related across domains. It would not be enough to have qualitative notions of these distinctions or similarities, since they must be able to be formatted in such a way, either via coding, language, or electronically, so as to be used by the machine. This book gives an interesting approach to the problem of concept classification, but it does so only from a qualitative point of view. It is a good start in this regard, and readers will gain a lot of insight into the problems that it addresses. It does not however give any advice on how to implement its ideas into a real thinking machine. Mathematical concepts are brought in order to talk more meaningfully about spaces of concepts, but they are really restricted to metric spaces and not general enough to deal with the plethora of concepts that could present themselves in typical environments. The book should be considered more as a work in philosophy, so those interested in this field might enjoy the book more than those who were expecting a book more geared towards artificial intelligence and computer science. Those readers interested in automated theorem proving or automated mathematical discovery might find the discussion on geometric categorization models of interest, and will find an interesting application of Voronoi tessellations, namely that of accounting for the varying sizes of concepts in a categorization. By far the most interesting chapter in the book is chapter 6, wherein the author gives a highly original discussion of inductive inference. The ability of human cognition to generalize from a limited number of observations is viewed (correctly) by the author as very impressive, but he is careful to note that inductive inference cannot be done free of side constraints. Quoting the philosopher J.S. Peirce and his evolutionary explanation of why induction is so effective, the author uses his theory of conceptual spaces to develop a theory of constraints for inductive inferences. The main notion in this theory is that of "projectability", which attempts to delineate the properties and concepts that are may be used in inductive inference. The author wants to arrive at a computational model of induction, and he offers interesting proposals for doing so, even if they lack immediate empirical justification. Central to the problem of induction the author argues is how observations are to be represented. This has been neglected in the history of philosophy he says, and so he then proceeds to outline his ideas on how to represent observations, distinguishing three levels, namely the 'symbolic', the 'conceptual', and the 'subconceptual.' At the symbolic level, observations are represented by describing them in a specified language. At the conceptual level, observations are characterized relative to a conceptual space. At this level induction is viewed as concept formation. At the subconceptual level observations are characterized by inputs from sensory receptors. Induction is then viewed as the attaining of connections between various inputs. The author views the processing taking place in artificial neural networks as an example of modeling at the subconceptual level. The problem of induction is more complicated than is typically presented in the literature, the author argues. Inductive inference will look different depending on which approach to observations is taken. In his elaborations on the processes of induction, one of the key issues that arises is the how discovery takes place across different domains. The process of conceptualizing across different domains takes place, as expected, at the subconceptual and conceptual levels. The symbolic level is delegated to formulating laws.
Rating:
Summary: An eye opener
Review: Profound piece of work. I am not a cognitive scientist, and this book is a bit technical, but it is still within reach of the motivated lay person. Gardenfors puts forward a a model to explain cognition that he calls "conceptual spaces." These conceptual spaces are at a level of abstraction in between the symbolic (used by AI types) and connectionist (Neural Nets). But what makes his conceptual spaces interesting and plausible is the position he takes that in this conceptual space, most reasoning is done by evaluating the analog of a distance between two aspects of a perception. Or, we find things to be similar if they are "geometrically" (measurably) closer on some limited number of dimensional scales. This is easy to follow for things like colors, but he doesn't stop there. He goes on to describe how this explains a wide variety of perceptions, as well as how we form and reform categories and concepts, and shows how this informs semantics and the process of induction. My only criticism is that some of the illustratios would have been more powerful in color.
Rating:
Summary: A new model of thought
Review: Profound piece of work. I am not a cognitive scientist, and this book is a bit technical, but it is still within reach of the motivated lay person. Gardenfors puts forward a a model to explain cognition that he calls "conceptual spaces." These conceptual spaces are at a level of abstraction in between the symbolic (used by AI types) and connectionist (Neural Nets). But what makes his conceptual spaces interesting and plausible is the position he takes that in this conceptual space, most reasoning is done by evaluating the analog of a distance between two aspects of a perception. Or, we find things to be similar if they are "geometrically" (measurably) closer on some limited number of dimensional scales. This is easy to follow for things like colors, but he doesn't stop there. He goes on to describe how this explains a wide variety of perceptions, as well as how we form and reform categories and concepts, and shows how this informs semantics and the process of induction. My only criticism is that some of the illustratios would have been more powerful in color.
Rating:
Summary: Excellent! Conceptual Spaces make sense to me.
Review: The essence of conceptual spaces, as I understand it, is that we can define concepts as regions in conceptual spaces. A conceptual space is defined by axes representing qualities. For example, color spaces are conceptual spaces, as is the tasting combos of sweet, bitter, salty. Your choice of qualitative measures deeply affects how you understand the world. 'Spose reality is an infinitely dimensional, then we have lots of choices for axes. We simplify and correlate by using all that coordinate transformation and axis projection stuff from 3D graphics! Heck Gardenfors even uses Delauney Triangulation (or polyhedralization). Criterion P, page 71 A natural property is a convex region of a domain in a conceptual space. Criterion C, page 105 A natural concept is represented as a set of regions in a number of domains together with an assignments of salience weights to the domains and information about how the regions in the different domains are correlated. Concept Combination, page 122 The combination CD of two concepts C and D is determined by letting the regions for the domains of C, confined by D replace the values of the corresponding regions for D. (contrast class p. 119), for example the "stone lions" outside the NYC library. Six Tenets of Cognitive Semantics, page 160 i) Meaning is a conceptual structure in a cognitive system (not truth conditions in possible worlds)
ii) Conceptual Structure are embodied (meaning is not independent of perception or of bodily experience).
iii) Semantic elements are constructed from geometrical or topological structures (not symbols that can be composed according to some system of rules).
iv) Cognitive models are primarily image-schematic (not propositional). Image-schemas are transformed by metaphoric and metonymic operations (which are treated as exceptional features on the traditional views).
v) Semantics is primary to syntax and partly determines it (syntax cannot be described independently of semantics).
vi) Concepts show prototype effects (instead of showing the Aristotelian paradigm based on necessary and sufficient conditions). Process of Abstraction, page 191 - Start with a collection of things. Identify and quantify individual objects. The determine the clusters. Step three: abstract the clusters into dimensions. Simple! I especially liked the notion that a metaphor is taking the spatial relationship of a cluster of concepts in one domain and using them in a new domain to help understand the new domain.
<< 1 >>
| 
