Modeling Software Development Process Complexity

Abstract

Modern software systems are growing increasingly complex, requiring increased complexity of software and software development process (SDP). Most software complexity measurement approaches focus on software features such as code size, code defects, number of control paths, etc. However, software complexity measurement should not only focus on code features but on features that cover several aspects of SDP in order to have a more complete approach to software complexity. To implement this approach, an extensive literature review for identifying factors that contribute to the complexity of SDP was performed and seventeen complexity factors were identified. As there were indications that the identified factors were not independent from each other but there were interrelations between them, statistical methods for identifying the underlined relations and refining them were applied, resulting to the final set of measures used in the proposed model. Finally, the proposed model has been tested in five software projects and the results were evaluated.