Cost-priority cognizant regression testing

Abstract

Any change in software requires all the test cases, developed earlier, to be re-executed. This is done to make sure that the changes have not affected the proper working of the software. Since it is usually impossible to re-run all the test cases, the number of test cases, to be reexecuted, must be reduced. Prioritization, reduction and selection are the three techniques used to accomplish the same. The basis of the prioritization of test cases can be the time of execution or the cost. A test case suite has many test cases. The sequence of test cases in a suite also plays a part in evaluating the effectiveness of a suite. Both cost and priority should be taken into account to determine the order of test suites. This work proposes a fitness function, which depends on both factors. This fitness function would decide the priority of the test cases. Since it is an optimization problem, Genetic Algorithms have been used to accomplish the task. In order to verify the technique, the test data of four testing cycles of a professional Enterprise Resource Planning System has been used.