An ontology-mediated validation process of software models

Abstract When errors in software modelling activities propagate to later phases of software development lifecycle, they become costlier to fix and lower the qual-ity of the final product. Early validation of software models can prevent rework and incorrect development non-compliant with client’s specification. In this paper we advocate the use of ontologies to validate and improve the quality of software models as they are being developed, at the same time bridging the traditional gap between developers and clients. We propose a general ontology-mediated process to validate software models that can be adapted in a broad range of software de-velopment projects. We illustrate this for Multi-Agent Systems (MAS) develop-ment providing early evidence of the soundness of our approach. We successfully validate and improve the quality of MAS models for a real-life development pro-ject, illustrating the ontology-mediated models validation in a commercial setting.