ABSTRACT
This chapter looks at two simple models of recognition and recall and how a formal comparison of these simple models has led to interesting and unexpected insights. It presents four classes of theories: network models of memory; global matching models; parallel distributed model; and dual process models. The chapter examines two relatively simple models of memory. The first theory is the threshold model of recognition, followed by the generate–recognize model of recall. The generate–recognize model is an explanation of how recall and recognition may be related to one another. The chapter considers two multiple-trace models, search of associative memory and MINERVA, and also examines two distributed storage models, theory of distributed associative memory and composite holographic associative retrieval model. In semantic networks, information is generally stored in one location in the network using a cognitive economy. Knowledge is represented in latent semantic analysis as the relations among concepts in the high-dimensional semantic space.
