Diagrammatic knowledge representation

Diagrams are used to facilitate problem solving in engineering, physics, geology, and other scientific areas. Diagrams store related elements adjacents to each other. For instance, it can be easily seen from a map that Iowa and Illinois are adjacent states, since they are placed next to each other. A nondiagrammatic or sentential representation of the map would have this information explicitly coded. A sequential search of the sentential representation would lead to the conclusion that Iowa and Illinois are adjacent states. Typical problem representations in artificial intelligence (AI) applications require vast amounts of storage, and problem processing requires extensive time consuming searches through knowledge bases. It is shown that if adjacency properties of a diagram are captured using a suitable data structure, then the search effort required to reach a valid conclusion is reduced. Also, the storage requirements of a diagrammatic representation are less than that of an equivalent sentential representation