WordNet is an on-line lexical reference system whose design is inspired by current

MacWhinney (1978) presented a computational model of the acquisition of morphophonology. The present chapter attempts to extend the model presented in that earlier paper to the acquisition of word-order patterns. This extension is supported by an examination of the previous research on syntactic acquisition. In the final section of the chapter, further possible extensions to phonology and semantics are considered. The crucial claim underlying the basic approach to both morphophonology and syntax is that use of a given rule system is governed by a system of alternative strategies. Within such a multileveled model, alternative strategies can be compared in terms of their relative complexity. In the present chapter, these alternative strategies are evaluated through application of the following analytic technique: 1. A relatively simple strategy that can account for at least some of the observed data is presented. 2. It is shown that there are at least some data that are best explained by this strategy. 3. It is shown that, at some point in development, the child produces forms that cannot be explained by this simple strategy alone. 4. A strategy of somewhat greater complexity and power is introduced and it is shown that this strategy can account for at least some of the data not explained by the simpler (and weaker) strategy. This line of argumentation proceeds until evidence has been presented for six alternative strategies in word-order processing.

If people represent concepts with perceptual simulations, two predictions follow in the property verification task (e.g., Is face a property of GORILLA?). First, perceptual variables such as property size should predict the performance of neutral subjects, because these variables determine the ease of processing properties in perceptual simulations (i.e., perceptual effort). Second, uninstructed neutral subjects should spontaneously construct simulations to verify properties and therefore perform similarly to imagery subjects asked explicitly to use images (i.e., instructional equivalence). As predicted, neutral subjects exhibited both perceptual effort and instructional equivalence, consistent with the assumption that they construct perceptual simulations spontaneously to verify properties. Notably, however, this pattern occurred only when highly associated false properties prevented the use of a word association strategy. In other conditions that used unassociated false properties, the associative strength between concept and property words became a diagnostic cue for true versus false responses, so that associative strength became a better predictor of verification than simulation. This pattern indicates that conceptual tasks engender mixtures of simulation and word association, and that researchers must deter word association strategies when the goal is to assess conceptual knowledge. Researchers increasingly report that modality-specific

The ability to categorize and use concepts e#ectively is a basic requirementofany intelligent actor. The utility-based approach to categorization is founded on the thesis that categorization is fundamentally in service of action, i.e., the choice of concepts made by an actor is critical to its choice of appropriate actions. This is in contrast to classical and similarity-based approaches which seek logical completeness in concept description with respect to sensory data rather than action-oriented e#ectiveness. Utility-based categorization is normative and not descriptive. It prescribes howanintelligent agent ought to conceptualize to act e#ectively. It provides ideals for categorization, speci#es criteria for the design of e#ective computational agents, and provides a model of ideal competence. A decision-theoretic framework for utilitybased categorization whichinvolves reasoning about alternative categorization models of varying levels of abstraction is proposed. Categorization mode...

A computational psycholinguistic model of written language comprehension called PM (Propositional Model) is described. PM is a highly interactive model. Written English text is processed directly into propositional representations. There is no separate syntactic analysis and no distinctly syntactic representations exist. The processing mechanism is lexically driven and most knowledge of language is assumed to be encoded in the lexicon. Of particular importance to the processing mechanism are the relational lexical items in the input text. These lexical items set up expectations which drive the processing mechanism and determine the possible propositional structures. PM's

How can an intelligent agent learn an effective representation of its world? This dis-sertation applies the psychological principle of cognitive economy to the problem of representation in reinforcement learning. Psychologists have shown that humans cope with difficult tasks by simplifying the task domain, focusing on relevant features and generalizing over states of the world which are “the same ” with respect to the task. This dissertation defines a principled set of requirements for representations in rein-forcement learning, by applying these principles of cognitive economy to the agent’s need to choose the correct actions in its task. The dissertation formalizes the principle of cognitive economy into algorithmic crite-ria for feature extraction in reinforcement learning. To do this, it develops mathematical definitions of feature importance, sound decisions, state compatibility, and necessary distinctions, in terms of the rewards expected by the agent in the task. The analysis shows how the representation determines the apparent values of the agent’s actions, and proves that the state compatibility criteria presented here result in representations

Abstract. Unsupervised organization of a set of lexical concepts that captures common-sense knowledge inducting meaningful partitioning of data is described. Projection of data on principal components allow for identification of clusters with wide margins, and the procedure is recursively repeated within each cluster. Application of this idea to a simple dataset describing animals created hierarchical partitioning with each clusters related to a set of features that have commonsense interpretation. Keywords: hierarchical clustering, spectral analysis, PCA. 1

This paper discusses stylistic variants for Verb Phrase(VP), V ’ and V anaphors in English and Japanese and finds some common features between those English and Japanese anaphors by dealing with the hierarchy of stylistic preferability among (i) expressions with full anaphors, (ii) those with zero anaphors where the ellipsis occurs with the auxiliaries like do and desu and (iii) those with lexical anaphors like do so, do it/do that in English and their corresponding ones in Japanese, and by observing in which style the anaphors occur or in which situations they often appear pragmatically. As a result, in both English and Japanese, I propose four types of stylistic hierarchy for the anaphors according to contrastive elements from different types of sentence construction. The hierarchies in four types are quite similar in both languages except that repeated verb anaphors which are related to cognitive identification are considered one of the best expressions in Japanese. They are proved by the examination of the frequency of the anaphors in use in spoken style, informal written style and in formal written style. Each of the characteristics of zero, lexical and full anaphors of both languages (as well as repeated verb anaphors in Japanese) reflects on their frequency of VP, V ’ and V anaphors in the three styles. _____________________________________________________________________ 1.

processes that detect tterns (e.g., lines, angles, closed areas various sorts, etc.), and relationships tween patterns, in this display. This rt of classification may utilize a series "procedures" (in Winograds 1973, 1975, nse) that test for the criteria associated [th membership in given conceptual tegories (like "pointedness," or "leg"). e present claim is that the same ,ocedures may be used appropriately for [assifying both internal representations -ising during perception which engender perience of a visual percept, and internal presentations experienced as a visual ntal image. wish to thank Steve Palmer and Jim merantz for helpful criticism and ggestions. Some of the work described rein was accomplished with the aid of [omedical Sciences Grant 5 SO5 RRO7041-09 arded to the Johns Hopkins University by he Division of Research Resources, DHEW. i46 An i.mage space: The psychological analogue to the CRT in our metaphor can be thought of as having spatial boundaries, although not

Double R Grammar is a linguistic theory of the grammatical encoding and integration of referential and relational meaning in English. Double R Grammar is fundamentally a Cognitive Linguistic theory (Langacker, 1987, 1991; Lakoff, 1988; Talmy, 2003) and the use of the term " grammar" encompasses both meaning and structure as it does in Cognitive Grammar (also known as Cognitive Semantics). Grammar is the symbolization of meaning. Discussion of grammar goes hand in hand with discussion of the meaningful consequences of grammatical variation. The linear stream of English text encodes multiple dimensions of meaning. The encoding of multiple dimensions of meaning in a single linear dimension results in trade-offs in encoding across dimensions and variation within a given dimension across different grammatical contexts. A primary goal of Double R Grammar is to provide integrated representations of referential and relational meaning that reflect these trade-offs.