期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2019
卷号：97
期号：4
页码：1205-1217
出版社：Journal of Theoretical and Applied
摘要：Identification of software components is found to be a crucial task and remains as a challenging area in the software domain to extract optimal components from the component repository[18]. Several methods are deployed to identify the software components and observed that clustering based technique is frequently used and offer a solution with certain limitations such as prior specification of a set of clusters, overlapping and difficulty in selecting the correct distance metric. In this research, an optimization technique is applied on some components to get the best result. The genetic algorithm using the concept of number of chromosomes is applied on software complexity metrics such as Cohesion of Variables within a Component (COVC), Cohesion of Methods within a Component (COMC) and Total Cohesion Complexity of a Component (TCCC) which are proposed by Rana and Singh et.al [14]. Fitness function metrics is proposed for finding out the fitness value. Rana and Singh also empirically evaluate cohesion metrics(COVC, COMC, TCCC) and perform comparative analysis with existing metrics in another research paper [15]. In this paper genetic algorithm is applied on these metrics for optimization of results. After application of genetic algorithm, SPSS a statistic tool is applied on the result to find out the significance of result. The result obtained shows that better optimization of software metrics is obtained through application of genetic algorithm compared to without usage of genetic algorithm.
