Combined Goal Recursion Strategy and Visual Perception Strategy on Human Problem Solving and Cognitive System Simulation

This paper mainly discussed cognitive strategy application in human problem solving through mind simulation combined brain-imaging experiment. An artificial intelligent cognitive system based on ACT-R (Adaptive Control of Thought-Rational) theory was built to simulate simplified 4 × 4 Sudoku problems. An fMRI (functional Magnetic Resonance Imaging) experiment was preformed to collect real data on brain activation pattern and to validate the system. It was assumed that participants used goal recursion and visual perception simultaneity when solving problem. Procedural rules integrated both strategies were introduced. Based on the assumption and procedural knowledge, the ACT-R cognitive system simulated both behavioral processing and brain activations. Compared ACT-R predictions and fMRI results, difference on response time is 0.1 and the average correlation on brain activation pattern of BOLD (Blood Oxygen Level-Dependent) response is about 0.9. The cooperation of goal module and visual module is illustrated to confirm the combination strategy. This study shows that combination of goal recursive strategy and visual perceptional strategy is effective in speeding up human problem solving. This study may shed light on the research of human-computer interface on intelligent system.