1. Introduction

The knowledge of the likely problematic areas of a software system is very useful for improving its overall quality. Based on such information, a more focussed software testing and inspection plan can be devised. More specifically, the limited resources allocated for software quality and reliability improvement can be expended in a cost-effective manner. Some of the practical software quality and reliability improvement techniques include, rigorous code design and code reviews, extensive software testing, and skill-based placement of personnel. In software development practice, the amount of project resources allocated for software quality improvement is usually a small fraction of the total budget, thus, asserting the importance of a cost-effective software quality improvement for allowing greater return on investment.

Software measurements, such as software product and process metrics, have been shown to be excellent indicators of software quality.1 ^,2 Based on such metrics, software quality classification (SQC) models can be built to predict the risk-based class membership of a software (program) module.3 For example, program modules can be predicted as either fault-prone ( fp) or not fault-prone ( nfp). With the aid of a SQC model, the software quality team can target the available resources to improve the modules predicted as fp, thus, allowing for a better resource utilization. A SQC model is built based on...