Empirical Investigation of Metrics for Fault Prediction on Object-Oriented Software

The importance of software-quality classification models which can predict the modules to be faulty, or not, based on certain software product metrics has increased. Such predictions can be used to target improvement efforts to those modules that need it the most. The application of metrics to build models can assist to focus quality improvement efforts to modules that are likely to be faulty during operations, thereby cost-effectively utilizing the software quality testing and enhancement resources. In the present study we have investigated the relationship between OO metrics and the detection of the faults in the object-oriented software. Fault prediction models are made and validated using regression methods for detecting faulty classes and discover the number of faults in each class. The univariate and multivariate logistic regression models are made by taking the dependent variable as the presence of fault or not. While linear regression models are built using the number of faults as dependent variable. The results of the two models are compared and an investigation on the metrics is presented.