Typed Abstract State Machines for data-intensive applications

摘要

This paper focuses on extensions of Abstract State Machines (ASMs) for the benefit of a refinement calculus in the area of data-intensive applications, in particular data warehouses and on-line analytical processing. The extensions lead to a typed version of ASMs and a more specific notion of strong data refinement, which incorporates the preservation of information content by means of schema dominance. Each typed ASM can be effectively translated into an equivalent ordinary one. Providing typed ASMs helps to exploit the existing logical formalisms used in data-intensive applications to define a ground model. Furthermore, strong data refinement helps to set up a refinement-based development method that uses provably correct standard refinement rules.