A method for predicting open source software residual defects

摘要

Nowadays many commercial projects use open source applications or components (OSS). A recurring problem is therefore the selection of the most appropriate OSS for a project. A relevant criterion for selection is the reliability of the OSS. In this paper, we propose a method that selects the software reliability growth model (SRGM), which among several alternative models best predicts the reliability of the OSS, in terms of residual defects. Several methods exist for predicting residual defects in software, and a widely used method is SRGM. SRGM has underlying assumptions, which are often violated in practice, but empirical evidence has shown that many models are quite robust despite these assumption violations. However, within the SRGM family, many models are available, and it is often difficult to know which models are better to apply in a given context. We present an empirical method that applies various SRGMs iteratively on OSS defect data and selects the model which best predicts the residual defects of the OSS. We empirically validate the method by applying it to defect data collected from 21 different releases of 7 OSS projects. The results show that the method helps in selecting the best model among several alternative models. The method selects the best model 17 times out of 21. In the remaining 4, it selects the second best model.