Using mutual information to select test suites in a black-box framework

Mutual Information is an information theoretic measure designed to quantify the amount of similarity between two random variables ranging over two sets. In this paper, we adapt this concept and show how it can be used to select a good test suite, in a black-box scenario and following a maximize diversity approach. We provide experimental evidence to show the usefulness of the measure. We also show that the time needed to compute the measure is negligible when compared to the time needed to apply extra testing. Finally, we compare our measure with current test prioritization measures and show that our proposal outperforms them. As a side result, in this thesis we present a Genetic Programming approach, fully supported by a tool, to generate test suites using Information Theory based measures.