Knowledge Representation using tne Two-Stage Entity-Relationship Approachi A Case Study in manufacturing Information Management

A metadatabase approach to data and knowledge management in Computer-Integrated Manufacturing is developed at RPI [Hsu 1988). Prototyping of the metadatabase has been underway since 1937; preliminary results include a global data model and some interfacing procedures for integrating an MRP II system and a shop floor control system [Hsu, et. al. 1987a]. This paper presents a rule base modeling methodology applied to the knowledge components of the metadatabase. The methodology itself is based on the Two-Stage Entity-Relationship (TSER) approach [Hsu 1985; Hsu et. al. 1987b] to information modeling. Previous results of TSER are extended to also covering production rules. In particular, the Operational Entity-Relationship construct is employed to model the logical interrelationships among rules. Current research on this methodology is focused on establishing a rigorous linkage between data representations and knowledge representations through TSER, thereby producing an integrated model for both data and their contextual knowledge.