Using fuzzy cognitive maps for data abstraction and synthesis in decision making

The abstraction of data into more usable formats is an important consideration in complex information processing environments. To be effective data abstraction must preserve the original behavior of the system it is modeling, while at the same time reducing the cognitive loading of the decision maker by reducing the volume of information he or she must process. Fuzzy cognitive maps can be used as a template for data abstraction. Because they represent the causal reasoning of the decision maker about a problem, they inherently incorporate how the decision makers use information. A heuristic is presented that preserves the input-output relationship of the map by preserving the cycles present in the map.