An Analysis of ACME: The Analogical Constraint Mapping Engine

This paper discusses some pragmatic issues on the analogical constraint mapping engine (ACME), a widely used artificial neural network model for analogical matching employing Grossbergs IAC (interactive Activation and Competition) artificial neural network model. Our analysis beings with an investigation into the use of a Hopfield constraint satisfaction network for image reconstruction. This domain providing us with a constraint satisfaction network based on the more usual Hopfield model, but which shares many characteristics with the ACME model. Based on this comparison we demonstrate how the convergence time increases exponentially with increasing network size, a factor which has serious implications for the practical application of this model to all but trivially small metaphors. Finally, we briefly present a localist connectionist model for solution generation in one domain, with linear scaleability over problem size.