A Systematic Mapping and Review Study on Software Defect Density Prediction Models

Defect density is a widely used metric in Software Engineering for evaluating the quality and reliability of software systems. Accurate prediction of defect density supports informed decision-making in software development and maintenance. Bio-inspired algorithms have demonstrated significant potential in solving optimization problems, including those related to software quality prediction. This study presents a systematic mapping and literature review focused on defect density prediction using bio-inspired algorithms. A total of 42 primary studies were selected through a rigorous search and filtering process conducted across major scientific databases. The results identify key trends, commonly used bio-inspired techniques, dataset sources, evaluation metrics, and hyperparameter optimization strategies. This review highlights existing research gaps and proposes future directions, including the integration of explainable Artificial Intelligence and hybrid metaheuristic approaches.