Information System Model and Architecture Generator

The advent of integrated, shared-data systems has made it increasingly necessary to address the application development process from the architectural and manufacturing perspective rather than from a build-as-you-go job shop viewpoint. Although the Business Systems Planning (BSP) methodology provides an enterprise wide strategic Information Systems plan, it is still at an abstraction level that leaves the traditional gap between " requirements " and implementations untouched. The Information System Model and Architecture Generator (ISMOD) tool complements and enhances BSP by mechanizing the planning process, thus providing a facility to narrow this gap by allowing orderly and consistent top-to-bottom architectural decomposition of the enterprise environment. It is an enterprise planning vehicle and not an implementation system, but it is the first critical component to support an integrated systems architecture effort. It automates and, to a large extent, formalizes a laborious require ments documentation process preceding code development , and it does this " top to bottom, " from a global, enterprise-wide, information requirements viewpoint. This paper discusses the overall architectural concepts of integrated data systems development, the place of ISMOD within it, and the specific facilities, techniques, and information provided by the system. T he Information System Model and Architecture Generator (ISMOD) was developed to aid enterprise executives and integrated data systems architects in the analysis of the information needs of an enterprise. Its primary purpose is the identification of the enterprise processes (functions) and data required to manage the resources of an enterprise. On the basis of data usage in the enterprise, the system will allow knowledgeable information systems architects to group various processes into process clusters called subsystems. Through a simulation capability, it is possible to project the impact a newly implemented subsystem may be able to achieve in terms of user satisfaction. Traditionally, the data processing industry has developed applications by targeting operational, high-volume systems, with the primary concern being efficient machine utilization. A very specific and rigid development methodology has been devised which is very successful in implementing these " bottom-up, stand-alone " systems. However, with the emergence of data base technology , many organizations have shifted their emphasis from stand-alone operational systems to integrated, shared-data systems. In this environment, it is extremely important that applications be viewed in the context of the whole enterprise. As a result, the development methodology used to create stand-alone applications is not appropriate, and in many cases, it is counterproductive. The development …