USING ACTIVATION NETWORKS FOR ANALOGICAL ORDERING OF CONSIDERATION: ONE METHOD FOR INTEGRATING CONNECTIONIST AND SYMBOLIC PROCESSING.

It is shown how activation networks can be used for analogical reasoning in the task of ordering the consideration of alternatives. An approach for the integration of the use of an activation network with symbolic processing is described. A prerequisite for this type of integration, as well as for allowing an activation network to use knowledge acquired for or by a knowledge-based system with symbolic processing, is to define a mapping between knowledge of the two types. The authors discuss analogical reasoning, the network structure, its processing, its integration with symbolic processing, learning, and a form of mixed reasoning.