Fuzzy Measure Extraction for Software Quality Assessment as a Multi-Criteria Decision-Making Problem

Being able to assess software quality is essential as software is ubiquitous in every aspect of our day-to-day lives. In this paper, we rely on existing research and metrics for defining software quality and propose a way to automatically assess software quality based on these metrics. In particular, we show that the software quality assessment problem can be viewed as a multi-criteria decision-making (MCDM) problem. In Multi-Criteria Decision Making (MCDM), decisions are based on several criteria that are usually conflicting and non-homogenously satisfied. Non-additive (fuzzy) measures along with the Choquet integral can be used: they model and aggregate the levels of satisfaction of these criteria by considering their relationships. However, in practice, it is difficult to identify such fuzzy measures. An automated process is necessary and can be used when sample data is available. We propose to automatically assess software by modeling experts’ decision process: to do this we automatically extract the corresponding fuzzy measure from samples of the target experts’ decision. We were able to improve previous approaches to automatic software quality assessment that used machine learning techniques.