EXPERT SYSTEM USING DEEP KNOWLEDGE AND SHALLOW KNOWLEDGE

An expert system infers by using deep knowledge and shallow knowledge. The deep knowledge may be physical rules which are basically satisfied in an electrical circuit which is an object of inference, and the shallow knowledge may be assumption knowledge derived from expert experience relating to the electrical circuit. The deep knowledge includes structure information of the object of inference and equation information representing behavior of the object. The structure of the shallow knowledge, which includes a concluding equation derived by omitting a portion of the equation of the deep knowledge, a conditional equation representing a condition under which the concluding equation is satisfied and the omitted item, is made essentially similar to the structure of the deep knowledge so that both the deep knowledge and the shallow knowledge are uniformly handled in one inference process.