Architecture level prediction of software maintenance

Abstract

A method for the prediction of software maintainability during software architecture design is presented. The method takes: the requirement specification; the design of the architecture; expertise from software engineers; and, possibly, historical data as input and generates a prediction of the average effort for a maintenance task. Scenarios are used by the method to concretize the maintainability requirements and to analyze the architecture for the prediction of the maintainability. The method is formulated based on extensive experience in software architecture design and detailed design and exemplified using the design of software architecture for a haemo dialysis machine. Experiments for evaluation and validation of the method are ongoing and future work.