Smith, E. and Medin, D. 1981. Categories and Concepts. Harvard University Press.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....needed a distinct name, our language would be staggeringly complex and communication virtually impossible. Fortunately, though, we do not perceive, remember, and talk about each object and event as unique, but rather as an instance of a class of concept that we already know something about [14]. This sort of argument makes a compelling appeal to intuition but it does not provide an answer in terms of any existing theories. In contrast, the present paper provides a an information theoretic account of the processes of and effects achieved by concept learning. The account shows that ....

Smith, E. and Medin, D. (1981). Categories and Concepts. Cambridge, Mass.: Harvard University Press.

....have pointed ears and long noses. Snowman heads are approximately round with carrot like noses. However, conceptual and perceptual categorization in humans are not yet well understood, and no one knows what information is used and what kind of processing takes place when constructing categories [19]. For these reasons, learning classes of objects for recognition purposes generally occurs in a supervised setting; that is, the class is learned from examples. This brings about a second difficulty. Learning from a finite set of examples is an ill defined problem that must be regularized by ....

E. Smith, L. Medin, Categories and Concepts. Cambridge, MA:Harvard University Press, 1981.

....situations even slightly outside the realm of those encoded in the rules fail completely, and the system exhibits discontinuous behavior) We see no way to repair this property of such systems. See also and 59 2. 2 Match with Psychological Results Psychological research on categorization ]34] 25] 13] 2] has shown that category formation cannot be explained in a,classical logical model. That is, the conditions of category membership are not merely logical conditions (result of expressions with connectives and or and not ) Rather, categories are organized around focus concepts ....

Smith, E.E., and Medin, D. Categories and Concepts. Cambridge, MA: Harvard University Press, 1981.

....this is the case when there is a single, obvious way in which a specific concept instantiates a general one, and when all the general properties of concepts in the general theory transfer to the domain specific ones. Psychological work on Category Structures provides support for this claim [Smith and Medin, 1981], and, computationally, in heritance is simple to understand and use. However, as we have argued in Section 1, inheritance becomes inappropriate when a general theory can be applied in multiple ways, and when we wish to restrict how and which properties transfer to more domainspecific concepts. ....

Smith, E. E. and Medin, D. L. (1981). Categories and Concepts. Harvard Univ., Cambridge, Ma.

.... associated with the concept (Barsalou [5] Linkages from Jkatures to concepts: Once all or most of the features of a concept are activated (implicitly or explicitly) the concept itself should be activated to cover these features (roughly the categorization process; Smith and Medin [41 ]) These four criteria strongly favor a two level architecture, with one level for explicit conceptual representation and reasoning and the other level for features and implicit, subconceptual processes. Such a framework (containing both types of reasoning) enables the exploration of their ....

E. Smith and D. Medin, Categories and Concepts. (Harvard University Press, Cambridge, MA, 1981).

....Geometric Model of Concept Formation Peter G ARDENFORS Cognitive Science, Lurid University, Kungshuset, Lundaggrd, S 223 50 LUND, Sweden 1. Introduction The classical account of concepts within philosophy is Aristotle s theory of necessary and sufficient conditions (see [19] for a presentation of this and other theories of concept formation) His view on how concepts are determined has had an enormous influence throughout the history of philosophy. During this century, the Aristotelian notions became part of the program of the logical positivists who demanded that ....

....representative members of a category are called prototypical members. It is well known that many properties, like red and bald have no sharp boundaries, and for these it is perhaps not surprising that one finds prototypical effects. However, such 1See e.g. 9] 2 See e.g. 17] 18] 14] [19], and [10] for extended discussions of the theory. effects have been found for most properties including those with comparatively clear boun daries like bird and chair . Now, if some form of the Aristotelian notion of a concept is adopted, it is very diffcult to explain these prototype ....

Smith, E. och Medin, D.L. (1981): Categories and Concepts', Harvard University Press, Cambridge, MA.

....In contrast, extrinsic features describe an entity relative to other entities and events. For example, used to strike nails is an extrinsic feature of a hammer because it relates a hammer to nails and striking. Efforts to represent concepts using only intrinsic features have largely failed [11], especially for representing artifacts [3] Although the distinction between intrinsic and extrinsic properties is more spectral than black and white, it is important to distinguish the many cases that fall into the uncontroversial extremes because they differ in significant ways. For example, ....

....inapplicable. Moreover, unlike intrinsic features, an entity s extrinsic features may be contradictory, such as the salary of a person with multiple jobs. For these reasons, most psychological research on concept representation distinguishes between an entity s extrinsic and intrinsic features [11]. From these distinctions (and others we discuss later) we draw three conclusions. First, the distinction between intrinsic and extrinsic features is important; a knowledge based system that ignores their differences might draw incorrect inferences. Second, the roles and purposes of an entity are ....

E. Smith and D. Medin. Categories and Concepts. Harvard University Press, Cambridge, MA, 1981.

....take into account the weights of instances and revising instance reassignment to generate class weights for every instance, which are then used to produce the next generation of class models. 3. 1 Prototype Modeler Our first supervised algorithm, which plays a role in Kmeans, creates a prototype [13] or centroid for each class by extracting the mean of each attribute from training cases for that class. Such a prototype modeler classifies an instance by selecting the class with the centroid closest to it in n dimensional space. Because the distance metric is sensitive to variations in scale, ....

E. E. Smith and D. L. Medin. Categories and Concepts. Harvard University Press, Cambridge, MA, 1981.

....In contrast, extrinsic features describe an entity relative to other entities and events. For example, used to strike nails is an extrinsic feature of a hammer because it relates a hammer to nails and striking. E orts to represent concepts using only intrinsic features have largely failed [9], especially for representing artifacts [5] It s important to distinguish between these two types of features because they di er in signi cant ways. For example, an entity s intrinsic features (such as age) may change over time, but its extrinsic features (such as the salary of a person) may ....

....inapplicable. Moreover, unlike intrinsic features, an entity s extrinsic features may be contradictory, such as the salary of a person with multiple jobs. For these reasons, most psychological research on concept representation distinguishes between an entity s extrinsic and intrinsic features [9]. From these distinctions (and others we discuss later) we draw three conclusions. First, the distinction between intrinsic and extrinsic features is important; a knowledgebased system that ignores their di erences might draw incorrect inferences. Second, the roles and purpose of an entity are ....

E. E. Smith & D. L. Medin. Categories and Concepts. Harvard University Press, Cambridge, MA, 1981.

....to take into account the weights of instances and revising instance reassignment to generate class weights for every instance, which are then used to produce the next generation of class models. 3. 1 Prototype Modeler Our first supervised algorithm, which plays a role K means, creates a prototype [13] or centroid for each class by extracting the mean of each attribute from training cases for that class. Such a prototype modeler classifies an instance by selecting the class with the centroid closest to it in n dimensional space. Because the distance metric is sensitive to variations in scale, ....

E. E. Smith and D. L. Medin. Categories and Concepts. Harvard University Press, Cambridge, MA, 1981.

....supports all of the higher cognitive functions. From memory to language to thought, this amodal data structure is manipulated independently of sensorymotor systems to represent the world, draw inferences, make decisions, and solve problems. For reviews of such theories, see E. Smith (1978) E. Smith and Medin (1981), Rumelhart and Norman (1988) Barsalou and Hale (1993) or any standard text in cognitive psychology or cognitive science. An alternative approach assumes that conceptual representations are grounded, at least to some extent, in sensory motor systems. According to these views, conception and ....

Smith, E.E., & Medin, D.L. (1981). Categories and concepts. Cambridge, MA: Harvard University Press.

....than relational content, insofar as it is concerned with content at all. For example, the psychological literature on concepts seems to be largely concerned with how concepts are applied to the actual world, concentrating on something like the primary intensions of the concepts involved. See Smith and Medin 1981 and Patterson 1991. 35 Because it is tied to language rather than thought, however, Kaplan s account has some marked differences with the account I have presented. On Kaplan s account, a name or a natural kind term has the same content in any context, so that its character is relatively ....

Smith, E.E. & Medin, D.L. 1981. Categories and Concepts. Cambridge, MA: Harvard University Press.

....[Michalski et al. 1983] and ID3 [Quinlan, 1986] KADS II TII.4.3 TR VUB 002 3.0 42 Machine Learning Issues in CommonKADS to classify examples of a concept. Such definitions may be adequate for mathematical and other strictly formal concepts, but as shown by several authors [Wittgenstein, 1953, Smith and Medin, 1981, Schank et al. 1986] it is extremely difficult to present classical definitions of naturally occurring concepts like game , bridge , chair , car , bird , etc. Smith and Medin [Smith and Medin, 1981] distinguish between the classical , the probabilistic (e.g. degree of prototypicality) ....

....and other strictly formal concepts, but as shown by several authors [Wittgenstein, 1953, Smith and Medin, 1981, Schank et al. 1986] it is extremely difficult to present classical definitions of naturally occurring concepts like game , bridge , chair , car , bird , etc. Smith and Medin [Smith and Medin, 1981] distinguish between the classical , the probabilistic (e.g. degree of prototypicality) and the exemplar (purely extensional) view of concept definitions. Non classical concepts are defined intensionally by a set of typical properties i.e. a set of default properties and or extensionally ....

Smith, E. and Medin, D. (1981). Categories and concepts. Harvard University Press, Cambridge, MA. KADS-II/TII.4.3/TR/VUB/002/3.0 50 Machine Learning Issues in CommonKADS

....properties 3 A central issue in cognitive science concerns how people store information about real world entities, and how this knowledge is accessed when a person reads or hears a word. Many theories assume that concepts are represented in terms of properties (for an extended discussion, see Smith Medin, 1981). For example, TIGER might be represented as an animal that has fur, has claws, has whiskers, and lives in the jungle. 1 However, a list of properties does not fully capture everything that people know about entities. For example, people also seem to encode knowledge of the relationships among an ....

Smith, E. E., & Medin, D. L. (1981). Categories and Concepts. Cambridge, MA: Harvard University Press.

No context found.

Smith, E. and Medin, D. 1981. Categories and Concepts. Harvard University Press.

No context found.

E. Smith and D. Medin. Categories and Concepts. Harvard University Press, Cambridge, MA, 1981.

No context found.

Smith, E.E., & Medin, D.L. (1981). Categories and concepts. Cambridge, MA: Harvard University Press.

No context found.

Edward E. Smith and Douglas L. Medin. Categories and Concepts. Cambridge, MA: Harvard University Press, 1981.

No context found.

Edward E. Smith and Douglas L. Medin. Categories and Concepts. Harvard University Press, Cambridge, MA, 1981.

No context found.

E.E. Smith and D.L. Medin. Categories and Concepts. Harvard University Press, 1981.

No context found.

Smith, E.E. and Medin, D.L. (1981) Categories and Concepts, Harvard University Press

No context found.

Edward E. Smith and Douglas L. Medin. Categories and Concepts. Cambridge, MA: Harvard University Press, 1981.

No context found.

E. E. Smith and D. L. Medin. Categories and Concepts. Harvard University Press, Cambridge, MA, 1981.

No context found.

E.E. Smith & D.L. Medin. Categories and Concepts, Harvard University Press, 1981.

No context found.

E.E. Smith and D.L. Medin. Categories and concepts. Harvard University Press, Cambridge, MA, 1981.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC