Automated software fault-proneness prediction based on fuzzy inference system

Abstract

The identification of a module's fault-proneness is very important for minimizing cost and improving the effectiveness of the software development process. How to obtain the correlation between inspection metrics and module's fault-proneness, hiding in the observed data, has been focused by very researches. In this paper, we propose the use of a fuzzy inference system for this purpose. In order to empirically evaluate the effectiveness of proposed approach, we apply it on empirical data published by Ebenau and NASA's Metrics Data Program data repository, respectively. Experiments results confirm that proposed approach is very effective for establishing relationship between inspection metrics and fault-proneness, and that its implementation don't require neither extra cost nor expert's knowledge, and it is completely automated. Novel approach can provide software project managers with reasonably suggestion and much-needed insights.