The Exchange of Engineering Information Using Complex Data Structures

The definition of the data structures required to enable the exchange of information defining complex manufactured products is the central problem addressed in this thesis. The emerging STEP standard demands the use of a common product model to support the exchange of information, and has defined its own data description language, called EXPRESS. This language is both readable by humans, and parseable by computers and is text based. The use of a graphical information modelling technique is an established first stage in the creation and definition of these common product models. The Nijssen Information Analysis Method (NIAM) is one technique used within the STEP standard project and the automatic generation of product models in EXPRESS by a computer, from a NIAM diagram drawn on a computer screen by the user, has been implemented. This thesis is concerned with the use of NIAM to elicit information from domain experts and the generation of usable EXPRESS schemas from the NIAM diagrams produced. The production of coherent product models is dependent on the coherency of the source NIAM diagram; some techniques for testing the coherency of a NIAM diagram are proposed, together with a means of testing the integrity of the semantics expressed by the user. It is also possible to automatically elucidate the type structure of an entity from the semantics expressed. The specification of an advanced NIAM to EXPRESS interface incorporating these technique is proposed