Software Quality Assessment Algorithm Based on Fuzzy Logic

Abstract

In this paper an attempt has been made to provide a new global evaluation approach of a specified software quality model extracted from a generic software quality model using an instantiation procedure. The evaluation is based on data extracted from an ambient distributed system composed of fusion and fission agents connected to input/output services. These data are linked to the appropriate metrics of our software quality model and we use quality factors stated in ISO standards and different models of researchers represented under an ontology. We use equivalent relations to link criteria that have the same meaning and fuzzy logic approach to evaluate the entire software quality model. Our work presents the following contributions: (i) creating a generic software quality model based on several existing software quality standards and formalized under ontology concepts (ii) proposing an instantiation algorithm to extract specified software quality model from a generic software quality models (iii) proposing a new global evaluation approach of the specified software quality model using two processes, the first one executes metrics related to sensors data and the second one uses the result of the first process using fuzzy logic approach evaluating the entire specified software quality model and end up with a final numerical result (iv) adding the variability of metric variables algorithm to determine the impact of a possible variation of one criterion on others and avoid their penalization. This can help to conduct a trade-off-analysis in the proposed quality evaluation approach.