An approach for the forecasting of the maintainability of system functionalities

Abstract

The programming of system functionalities requires the development of components that must cooperate to satisfy the functional requirements of users, modularity, and reusability. Therefore, maintenance tasks demand developers to understand their internals and have knowledge about the dependency graphs that are formed by the interaction between components. Furthermore, they should be aware of the size, complexity, and maintainability of individual items, as well as of their aggregated weight for the complete coupling graph. Consequently, this paper aims to present an approach for calculating a forecasting indicator of the maintainability of system functionalities using as a base the indirect coupling graphs and a set of computed metrics for a group of code commits performed during a time period. Therefore, its contributions are a set of metrics for the calculation of a Maintainability Index of system functionalities and the individual elements in their dependency graphs, and to forecast the maintainability of system functionalities based on the sum of weights of the methods in the dependency graphs.