Can Fuzzy Mathematics enrich the Assessment of Software Maintainability?

Software maintainability depends both on qualitative and quantitative data. Existing maintainability models aggregate data into hierarchies of characteristics with given dependencies. However, data used to score the characteristics can be uncertain or even completely unknown. Therefore, it would be meaningful to evaluate sensitivity of the aggregated result, i.e. the maintainability, with respect to the uncertainty and incompleteness of data. In addition, real cases require an aggregation model able to evaluate the impact of changing the dependencies among the characteristics in the hierarchy on the maintainability. In this paper we argue that fuzzy mathematics can help to solve these problems. In particular, we show how a fuzzy aggregation model can be adopted to evaluate maintainability according to a hierarchical model.