An engineering design database system that consists of a query interface, a knowledge representation structure called the semantic association graph (SAG), and a design store is presented. SAG classifies existing design specifications by exploiting the common capabilities of a set of designs. The SAG generation algorithm computes the subsumption relations among design specifications (if they exist). The creation, evolution, and querying of the SAG structure are presented. The designer simply states the input requirements in the form of a simple declarative query which is mapped into the SAG structure. The system then suggests a set of designs that partially match the input requirements. The designer then takes these suggestions and retrieves the detailed design knowledge from the design store. The resulting scheme resembles KL-ONE, but has additional properties such as node sharing to model overlapping concepts
Published in:
Systems Integration, 1990. Systems Integration '90., Proceedings of the First International Conference on
Date of Conference: 23-26 Apr 1990