Software Engineering Research

Software engineering research in information systems examines how large codebases evolve over time, asking questions like how developers can safely restructure existing code without introducing new bugs, how duplicated code fragments spread and degrade maintainability, and how static properties of source code can predict where defects are likely to emerge. Empirical studies in this space apply machine learning to historical project data—commit histories, bug reports, code metrics—to build models that guide decisions about testing priorities, refactoring opportunities, and the tracing of requirements through to implementation. A persistent challenge is that models trained on one project often fail to generalize to others, and ground-truth labels for phenomena like intentional code clones or valid refactoring moves are expensive and inconsistent to collect. Active research directions include improving cross-project defect prediction, automating the recovery of traceability links between requirements and code, and understanding how API usage patterns both reflect and constrain the evolution of software systems.

Works

108,348

Total citations

1,589,039

Keywords

RefactoringCode Clone DetectionSoftware Defect PredictionRequirements TraceabilityStatic Code AttributesMachine Learning