摘要

This paper proposes a fuzzy qualitative and quantitative softgoal interdependency graphs (FQQSIG) model for non-functional requirements (NFRs) correlations analysis in Trustworthy Software (TS), which is considered a critical issue by academia, government, and industry. First, the FQQSIG model constructs an NFRs criteria decomposition hierarchy graph in the complex TS situation. Subsequently, it draws on trapezoidal fuzzy number and introduces a simulation algorithm named RAGE (RAndom GEneration) to transform the qualitative degree of importance assessments on NFRs in the form of experts' fuzzy linguistic variables into quantitative values. Using a new algorithm named Relationship Matrix, this model can calculate all NFRs contribution values by which developers could make tradeoff decisions among NFRs competing alternatives. The credibility and advantages of the FQQSIG model are discussed and an example is given to illustrate the FQQSIG model.