Incompleteness in Conceptual Data Modelling

Abstract

Although conceptual data modelers can "get creative" when designing entities and relationships to meet business requirements, they are highly constrained by the business rules which determine the details of how the entities and relationships combine. Typically, there is a delay in realising which business rules might be relevant and a further delay in obtaining an authoritative statement of these rules. We identify circumstances under which viable database designs can be constructed from conceptual data models which are incomplete in the sense that they lack this "infrastructural" detail normally obtained from the business rules. As such detail becomes available, our approach allows the conceptual model to be incrementally refined so that each refinements can be associated with standard database refactorings, minimising the impact on database operations. Our incremental approach facilitates the implementation of the database earlier in the development cycle. © Springer-Verlag Berlin Heidelberg 2013.