It has been difficult to handle unexpected exceptions in collaborative design. The expert system based method is extensively used, but it seems mission impossible not to loss some information when storing the expert experience in structured format, and to interact friendly in natural language. Aiming at these, we proposed a knowledge acquisition based approach for unexpected exceptions handling in collaborative design, which enable design engineer to define an exception using natural language, and make expert experience stored in natural language which is zero information loss. In this paper, two methods for mining similar records are proposed, and the performance comparison is shown that our approach is more efficient, extendable and precise.
展开▼