This research develops methods for extending data modeling capabilities for database systems. The role of a conceptual data model is to store information in a manner that provides for efficient processing by the computer while allowing the human to access this data in as natural a manner as possible. The approach taken in this project is to automatically generate data models to fit the requirements of the application environment. Research issues include (1) The formal specification and representation of a family of semantically rich conceptual data models, (2) the development of a rule-based approach to the generation of of database systems, and (3) the characterization of the performance of systems generated by this approach. The prototype system produced is tested against previously defined data models and refined based upon the feedback gained. The significance of this research is that it offers an approach to removing the inflexibility of present day database systems. Improving database systems is critical to extending their use, especially for engineering and manufacturing information, enhancing computer assisted systems which are rapidly becoming the vital core of design and manufacturing.
