Home :: Business Solutions: Research & Development

Software Measurements: Issues & Solutions

Most new software development projects nowadays depend heavily on object-oriented technology. While many benefits have been claimed for the object-oriented paradigm, it has proven a difficult challenge to manage the quality of large OO applications in practice.

The main goal of this research is to identify object-oriented metrics that can predict fault-prone components. In particular, metrics that can be collected from design documents can provide early indications of where quality problems lie. This is contended to be meaningful because object-oriented metrics are believed to be indicators of psychological complexity, and classes that are more complex are likely to be faulty. It has been hypothesized that the structural properties of a software component (such as its coupling) have an impact on its cognitive complexity. Cognitive complexity is defined as the mental burden of the individuals who have to deal with the component, for example, the developers, testers, inspectors, and maintainers. High cognitive complexity leads to a component exhibiting undesirable external qualities, such as increased fault proneness and reduced maintainability.

This research consists of two elements. First, the empirical evaluation of statistical and machine learning techniques for predicting fault-proneness of object-oriented classes. Second, developing a better understanding of the functional relationship between object-oriented metrics and the incidence of faults. These two elements are complementary in that a better understanding of the functional form of the metrics to faults relationship can result in improvements in prediction models.