A Metrics Suite for Object-oriented Design to Improve Software Quality

Accurate prediction of fault prone modules in software development process enables effective discovery and identification of the defects. Specific models and algorithms referred to identify the fault proneness in a system. Such prediction models serve the large-scale systems, where verification experts need to focus on development process. The ability of software quality models to accurately identify critical components allows for the application of focused verification activities ranging from manual inspection to automated formal analysis methods. Software quality models, thus, help ensure the reliability of the delivered products. The basic hypothesis of software quality prediction is that a module currently under development is fault prone if a module with the similar product or process metrics in an earlier project (or release) developed in the same environment was fault prone. Therefore, the information available early is directly proportional.