A prioritization model for software FMEA

The traditional software failure mode effect analysis (SFMEA) only considers the severity of the failure effect as the decision reference of the implementing priorities of the improvement measures. However, there are two issues for this as follows: 1) The analysis results are not sufficient, as it does not consider the difficulties of implementing the improvement measures; 2) The failure modes in the same severity level cannot be distinguished. To solve these two issues, a new measure i.e. `alterability´, is proposed to evaluate the consuming resources and the influences for implementing the improvement measures. The evaluation method of the alterability is presented in form of the scoring table. This paper introduces the severity and the alterability into the risk matrix model to get the Borda ordered values of this risk matrix model as the decision reference of the implementing priorities of the improvement measures. The Borda ordered value that considers the failure effect and the implementing difficulties of the improvement measures comprehensively makes the decision reference more sufficient and makes the failure modes in the same severity level be distinguished due to the different Borda ordered values. In the SFMEA of a real certain project, this proposed approach is applied to evaluate the results of the FMEA and guide the developers to determine the implementing priorities of the improvement measures according the Borda ordered values. The experimental results show that this proposed approach is feasible and efficient.