Search-Based Approaches for Software Modularization/Re-modularization

Abstract

Search-based Software Engineering helps in providing efficient decision support, and well as certain insights as per the requirement. The theory of the concept is important to understand because these approaches could be applied to every SDLC phase (Software Development Life Cycle Model) from a specification requirement point of view to its maintenance as well as operations. Search issues like solving SE issues in search space implies to find an ideal solution. The complete distribution of techniques for the prediction of defects is detailed further. A total of 77.5% of the studies are done on different classification methods, where 1.5% is focussed on association and clustering methods, and lastly, 14% analysis and studies are done on the different estimation techniques. Additionally, nearly 35.21% of the conducted research makes use of private datasets, whereas 64.79% of the conducted studies make use of public datasets. To predict software defects, nineteen distinct techniques were implemented. Seven of the most used techniques are recognized in the forecast of software defects from the nineteen techniques. However, there does exist numerous researches detailing the use of the search-based method to predict the maintainability as well as the change in proneness. It was discovered that application results when the search-based methods were followed for defect prediction and effort estimation were highly encouraging. Hence the motivation for this study was based on which the associations of the results would help the researchers and practitioners to get a set of guidelines and it would help them in making better choices of using these search-based methods in its specific conditions.