Research Article
Application of Evolutionary Algorithms for Software Maintainability Prediction using Object-Oriented Metrics
@INPROCEEDINGS{10.4108/icst.bict.2014.258044, author={Ruchika Malhotra and Anuradha Chug}, title={Application of Evolutionary Algorithms for Software Maintainability Prediction using Object-Oriented Metrics}, proceedings={8th International Conference on Bio-inspired Information and Communications Technologies (formerly BIONETICS)}, publisher={ICST}, proceedings_a={BICT}, year={2015}, month={2}, keywords={empirical validation evolutionary algorithms object-oriented metrics prediction modeling and analysis software maintainability prediction}, doi={10.4108/icst.bict.2014.258044} }- Ruchika Malhotra
Anuradha Chug
Year: 2015
Application of Evolutionary Algorithms for Software Maintainability Prediction using Object-Oriented Metrics
BICT
ACM
DOI: 10.4108/icst.bict.2014.258044
Abstract
The cost incurred during maintenance phase of any software consists of nearly 60-70% of the total project cost. In order to control, it needs to be measured in the earlier phases of software development life cycle (SDLC). Software Maintainability Prediction (SMP) is desirable because firstly the resource planning can be optimized in advance and secondly it helps in producing cost effective software systems. Significance of the Evolutionary Algorithms (EA) has substantially increased in recent times due to their capability of maximizing the quality function. Inspired by the evolutionary algorithms, we have conducted an empirical study for exploring the application of the EA for SMP. Although several traditional methods such as statistical and machine learning were applied in past, we experimented to apply EA for the first time for SMP. Two open source software projects Apache Poi 3.9 and Apache Rave 0.21.1 written in Java languages were used to carry out this empirical investigation and the results were analyzed using prevalent prediction accuracy measures. We observed that the optimization values were achieved more accurately and precisely with EA than the traditional methods, thus can be successfully applied for SMP.
