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：2015
卷号：82
期号：3
出版社：Journal of Theoretical and Applied
摘要：Testing is the best way to ensure software quality. However, this activity makes a huge challenge about limited time and financial resources. In order to achieve high quality, managers need to detect the defect prone parts of code and allocate the resources to them. Although many metrics, techniques and models have been proposed, effective identification of such parts is still a challenge. In this paper two new metrics, namely PIEDG and PIMDG are proposed for predicting defects based on dependency between the file level components of the code. The proposed metrics use the notion of participation of nodes in the software dependency graph to recognize the parts of the code which possess defects. This research also investigates the effect of ego graph on the software dependency global graph. The feasibility of the software defect predictor which is built based on the proposed metrics is evaluated. The experiments over the Eclipse bug dataset demonstrate promising results in anticipation of the software defects.
