ACQUISITION OF MANIPULATIVE GROUNDED SYMBOLS FOR INTEGRATION OF SYMBOLIC PROCESSING AND STIMULUS-REACTION TYPE PARALLEL-PROCESSING

The architecture constructed with two types of processing, logical sym bol processing and stimulus-reaction type parallel processing, seems p romising for intelligent systems. Since symbol processing is construct ed by a top-down approach and stimulus-reaction type processing is bui lt up by a bottom-up approach, a discrepancy, which is called the 'sym bol grounding problem', takes place. This paper presents a framework f or integration of symbol processing and stimulus-reaction type process ing from the viewpoint of solving the symbol grounding problem. In thi s framework designers or users use the conventional heuristic symbols and the systems use the self-organized symbols based on the characteri stics/environment of the systems themselves. Translation from one to a nother produces the fusion of those two symbols. The self-organized sy mbols are grounded and manipulative. Navigation of an autonomous robot is simulated. Acquisition of manipulative,grounded symbols with the p roposed framework is demonstrated. Since the constructed robot is equi pped only with a stimulus-reaction type controller, it has a robustnes s against noise and temporary geometrical changes.