Metrics Threshold Analysis On the Basis of Clustering Technique for Fault Prediction

Software metrics encapsulates many software quality factors such as fault proneness, reusability, and maintenance time of the software code. Software metrics are important implements to audit software quality throughout the project lifecycle. Software metrics are numbers accumulated from software code to assess and evaluate where problems are more probable to occur. In this work it is proposed to use a clustering and metrics thresholds based software fault prediction approach and explore it on the dataset, accumulated from Promise Data Repository. Grey relational technique is taken into account as the source for normalized values. The results of the grey relational analysis are then used to conduct fault-proneness classification based on the accuracy (F-measure) of one dataset and compared against the results of another datasets. The result obtain in this work demonstrate the effectiveness of metrics thresholds. This work is validated when the fault labels are unavailable and there is a need to check the accuracy of the software.