In software development projects, bugs are usually accumulated and technical debt gets bigger over time. Managers decide to reduce the technical debt by planning one or more iterations for bug fixing. Thetime required to fix a bug depends on the required skill and the resource skill level. Managers seek to achieve fixing the highest number of bugs during the iteration while at the same time fixing the highest possible number of high severity and high priority bugs. In this study, we optimize the human resource assignment to achieve the objectives above, using multi-objective evolutionary algorithms, and then we add a fourth objective, i.e. that the bugs left out of the iteration should require the least time to finish. We show that the additional objective can be optimized without the detriment of other objectives. The lesson is that complicating the multi-objective problem formulation can help with the overall quality of the solutions.
展开▼
