Implementing a model-driven and iterative quality assessment life-cycle: a case study

Assessing software quality through quantitative and reliable information is a major concern of software engineering. However, software is a complex product involving interrelated models with different abstraction levels targeting different stakeholders and requiring specific quality assurance methods. As a result, although Software Quality has gained maturity from a theoretical point of view, the practical quality assessment of software still does not fulfil enough involved actors’ expectations. In order to improve quality assurance in practice, a more integrated approach to assessment is required. This paper describes a case study in which a quality assessment framework (MoCQA) relying on a model-driven and iterative methodology has been used to this end. For a year and a half, the framework has been used by the quality assurance team of a small IT department to maintain and monitor a portfolio of projects in both production and development. The study shows the feasibility and the relevance of a model-driven and iterative quality assessment methodology in a professional environment. Besides, although its results still require more generalisation, the study provides interesting insights on how such an approach may help ensure a continuous and explicit communication between stakeholders, leading to a more efficient quality assessment.