Prioritizing Software Components Risk: Towards a Machine Learning-based Approach

Technical Debt (TD) can be detected using different methods. TD is a metaphor that refers to short-term solutions in software development, which may affect the cost of the software development life-cycle. Several tools have been developed to detect, estimate, or manage TD. TD can be indicated through smells, code comments, and software metrics. Machine learning Techniques (MLTs) are used in many software engineering topics such as fault-proneness, bug severity, and code smell. In this paper we use four internal structure metrics to identify and classify Architecture Technical Debt (ATD) risk by using MLTs. We show that MLTs can identify and classify the risk of ATD on software components to help the decision-makers to prioritizing the refactoring decisions based on the level of the risk.