Models of Memory

摘要: Over the past 50 years, models for human memory have developed from simple data descriptions specific tasks to general frameworks that can and been generalized most of major paradigms are used in research. We argue such a generalization across becomes possible because specify basic mechanisms rather than equations describe task. illustrate this by brief review developments since 1950s, followed discussion theoretical over 25 years. The early as Estes’ Stimulus Sampling Theory focused on learning, but 1960s emphasis gradually shifted especially distinction between short-term long-term memory. In 1980s number global were dealt with variety tasks. Although these quite successful there remained some problems, notably explanation lack list-strength effects recognition. Recent show based Bayesian or rational approach (ACT-R, REM) may provide unified framework explicit well implicit memory. Keywords: associative networks; ACT; distributed models; global models; mathematical models; SAM; TODAM