A Holistic Developer-Oriented Model for Estimating the Effort of Maintenance Tasks

Software maintenance effort/cost estimation aims at providing project managers with information on the expected effort needed to complete a specific maintenance activity. Previous work investigated potential maintenance cost drivers and built prediction models that support project managers in estimating the likely effort needed to issue a new release of the system. However, we believe that the current state of the art suffers some important limitations due to (i) lack of generality, since only approaches working for specific maintenance activities (e.g., bug resolution) have been proposed, while none of them can estimate the effort of a generic maintenance task, (ii) a gap between the cost drivers identified and the independent variables used in the devised models, and (iii) context-dependent estimations that depend on programming language or external factors. Furthermore, existing effort estimation models do not consider developer-related factors whose potential for effort estimation has been empirically shown recently. In this paper, we aim at overcoming such limitations by proposing a novel effort estimation model that incorporates cost drivers missed by previous models, and in particular developer-related factors, and is able to predict the class cost of a generic maintenance task using a five-level Likert scale. The empirical evaluation---that was conducted on 17 software systems and led to the construction of a new effort estimation dataset---showed that the proposed model is pretty accurate in the estimations reaching up to 70% of F-Measure and 80% of AUC-ROC. Furthermore, it is statistically better than other baseline models in 88% of the cases.