Approximate Reasoning for Processing Uncertainty