From conceptual modeling to requirements engineering

The critical impact of requirements analysis on software quality has been recognized long and repeatedly. Recent surveys in Europe and the USA have confirmed the problem on a much larger scale. One in two IS project failures can be attributed to a lack of quality in the requirements document.

Improving the quality of this document, as well as the practice of Requirements Engineering (RE), is therefore a prime objective. This objective is not easy to achieve, given the broad spectrum of considerations to be covered, the multitude of processes and products involved, the multitude of players to be involved, and the variety of pitfalls to be avoided. IE is a discipline at the interface of software engineering and systems engineering, which aims to provide models, methods and tools for developing and evolving quality requirements documents. This article aims to provide an overview of recent developments in this relatively young discipline, and to explore some of them in greater depth. EI extends the classical approach of trying to understand what the system is supposed to do, by trying to understand the "why" of the system, its raison d'être.

The "why" is expressed in terms of organizational goals and their impact on the requirements imposed on the information system. The article focuses on this dimension, and proposes and illustrates, through a case study, a goal-directed requirements engineering approach.