A Decision-Support System Approach to Economics-Driven Modularity Evaluation

Modularity debt is the most difficult kind of technical debt to quantify and manage. Modularity decay, thus modularity debt, causes huge losses over time in terms of reduced ability to provide new functionality and fix bugs, operational failures, and even canceled projects. As modularity debt accumulates over time, software system managers are often faced with a challenging task of deciding when and whether to refactor, for example, choosing to improve modularity or not. While the costs of refactoring are significant and immediate, their benefits are largely invisible, intangible, and long term. Existing research lacks effective methods to quantify the costs and benefits of refactoring to support refactoring decision making. In this chapter, we present a decision-support system (DSS) approach to the modularity debt management. Using such a system, managers would be able to play out various "what-if" scenarios to make informed decisions regarding refactoring. Our DSS approach is built on a scientific foundation for explicitly manifesting the economic implications of software refactoring activities so that the costs and benefits of such activities can be understood, analyzed, and predicted. We discuss our contributions and current progress in developing the building blocks and the underpinning framework, an integrated economics-driven modularization evaluation framework, for the modularity debt management decision-support system (MDM-DSS).