EXTENSION OF THE TEMPORAL SYNCHRONY APPROACH TO DYNAMIC VARIABLE BINDING IN A CONNECTIONIST INFERENCE SYSTEM

The relationship between symbolism and connectionism has been one of t he major issues in recent artificial intelligence research. An increas ing number of researchers from each side have tried to adopt the desir able characteristics of the approach. A major open question in this fi eld is the extent to which a connectionist architecture can accommodat e basic concepts of symbolic inference, such as a dynamic variable bin ding mechanism and a rule and fact encoding mechanism involving nary p redicates. One of the current leaders in this area is the connectionis t rule-based system proposed by Shastri and Ajjanagadde. The paper dem onstrates that the mechanism for variable binding which they advocate is fundamentally limited, and it shows how a reinterpretation of the p rimitive components and corresponding modifications of their system ca n extend the range of inference which can be supported. Our extension hinges on the basic structural modification of the network components and further modifications of the rule and fact encoding mechanism. The se modifications allow the extended model to have more expressive powe r in dealing with symbolic knowledge as in the unification of terms ac ross many groups of unifying arguments.