Representing the Richness of Linguistic Structure in Models of Episodic Memory

Abstract

The principal aim of a cognitive model is to infer the process by which the human mind acts on some select set of environmental inputs such that it produces the observed set of behavioral outputs. In this endeavor, one of the central requirements is that the input to the model be represented as faithfully and accurately as possible. However, this is often easier said than done. In the study of recognition memory, for instance, words are the environmental input of choice—yet because words vary on many different dimensions, and because the problem of quantifying this variation has long been out of reach, modelers have tended to rely on idealized, randomly generated representations of their experimental stimuli. In this paper, we introduce new resources from large-scale text mining that may improve upon this practice, illustrating a simple method for deriving feature information directly from word pools and lists.


Back to Table of Contents