Rethinking cognitive architecture via graphical models

Abstract

Cognitive architectures need to resolve the diversity dilemma – i.e., to blend diversity and uniformity – in order to couple functionality and efficiency with minimality, integrability, extensibility and maintainability. Building diverse architectures upon a uniform implementation level of graphical models is an intriguing approach because of the homogeneous manner in which such models produce state-of-the-art algorithms spanning symbol, probability and signal processing. To explore this approach a hybrid (discrete and continuous) mixed (Boolean and Bayesian) variant of the Soar architecture is being implemented via graphical models. Initial steps reported here, including a graphical implementation of production match and the beginnings of a mixed decision cycle incorporating a simple semantic memory, begin to show the potential of such an approach for cognitive architecture.