A model-level mutation tool to support the assessment of the test case quality

Abstract

Although mutation testing is a well-known technique for assessing the quality of tests, there is not a lot of support available for model-level mutation analysis. It is also considered to be expensive due to: (i) the large number of mutants generated; ii) the time-consuming activity of determining equivalent mutants; and (iii) the mutant execution time. It should also be remembered that real software artefacts of appropriate size including real faults are hard to find and prepare appropriately. In this paper we propose a mutation tool to generate valid First Order Mutants (FOM) for Conceptual Schemas (CS) based on UML Class Diagrams and evaluate its effectiveness and efficiency in generating valid and non-equivalent mutants. Our main findings were: 1) FOM mutation operators can be automated to avoiding non-valid mutants (49.1%). 2) Fewer equivalent mutants were generated (7.2%) and 74.3% were reduced by analysing the CS static structure in six subject CSs.