Ranking Software Risks Based on Historical Data

摘要

Risk ranking is a key step for risk management. This study divides ranking methods of software risks into two steps. The first step initializes a risk set from historical data. The second step achieves metric values for the risks in the set as ranking references. Although historical data has gained prominence when getting initial risks for ranking software risks, the existing ranking methods only use results of the second step. The ranked values of the initial risks from the first step are not exploited. Actually, ranked values of historical data are sometimes beneficial. Another kind of ranking method by this study utilizes the metric values or positions from both steps (RHD-2). The method exploits historical data properly and gets suitable ranked results for decision-makers by utilizing proper empirical coefficients according to the practical situations.