Graphical user interface testing using evolutionary algorithms

In software applications industry, the most important challenge is to ensure the maximum quality for software products. The demand for Graphical user interface (GUI) testing has very much increased in the last years. An important progress in this field has been achieved, from GUI manual testing to complete automate GUI testing. In this paper, optimization theory and evolutionary algorithms (EA) concepts were applied for GUI testing. In our approach, the GUI is completely separated from the code itself. The aim of the proposed methodology is to find the test case sequence(s) which produce(s) the maximum amount of changes inside the GUI while preserving all other constraints. The method starts by randomly generating the initial set of test cases, which is then improved through measuring and evaluating specific fitness functions. By using optimization theory, very good results were obtained for accelerating the automatic test of large interfaces.