Enhancement ofregression testing using genetic data generation and test case prioritization using m-ACO technique

Abstract

The changes that occur during the software development process is rapid. Hence software has to undergo modification frequently. Due to this modification,the cost for testing increases due to repetitive retesting. This retesting process is called as the regression testing. Modification made in the single test case will make the side effect in all other related test cases. In order to overcome this problem all the test cases have to be retested again and again whenever the changes are incorporated in the software. But testing all the test cases is time consuming and will also increase the cost of testing. To address this problem, this work focuses on providing priority to the test cases. Test case which had more effect to changes is assigned with higher priority and the test case which had the less effect to changes is assigned lower priority. For test case prioritization, we employ m-ACO (Modified Ant colony optimization) method.Test case prioritization is done in two ways namely “Triangle classification problem” and “Quadratic Equation Problem”. Flow of the data in the test case is done by Genetic Algorithm. This identifies the changed code in the program under test. It identifies both indirectly and directly affected def-use association in the modified part of the software by using forward walk algorithm and backward walk algorithm.