Testing-based Model Learning Approach for Legacy Components

Operating, maintaining, and upgrading legacy systems are the foremost challenges which are being faced by many organizations today. Usually, these systems are based on outdated technologies, have limited documentation, and actual developers are unavailable. It is risky to upgrade black-box legacy systems without knowing their internal structures. In this paper, we have proposed an approach which is based on the state of the art dynamic analysis technique known as Model Learning, a reverse engineering approach, to infer the behavioral models of legacy systems. We prepared and utilized our test-bed for black-box vending machines (considered as legacy systems) to learn the behavioral models of all the software modules embedded in vending machines. The in-depth analysis of learned models is helpful in the operation, up-gradation, and maintenance of the legacy system. The experimental results reveal that our proposed approach is very auspicious to modernize the legacy components and explore the concealed structures of the black-box systems automatically.