TEST CASE MINIMIZATION APPROACH FOR RELIABLE REGRESSION TESTING FOR WEB-BASED SYSTEMS

Abstract

Software testing is performed to ensure the correct operations of any software application. Different software testing techniques have been presented so far for this purpose. Regression testing is used to ensure the correct functionality of a software after some changes due to requirements. Hence, to perform regression testing test cases are used to test the ripple effect. A rigorous testing may help in meeting the quality criteria. However, a reduced and prioritized set of test cases may help in meeting the minimum required quality criteria for timely delivery of the software application. In this research, an approach named as TestReduce has been presented to select a prioritized minimal set of test cases to ensure that the web application meets the minimum quality criteria. TestReduce is proposed based on genetic algorithm to minimize regression test cases. The purpose of TestReduce is to optimize a solution in the form of test cases that may best fit to find out a specific set of test cases to initiate regression testing. The selected set will help to test the quality of the corrected code and to analyse the ripple effect. The ultimate goal of this research is to provide a way that may solve the prioritization problem of regression test cases in the case of linked requirements. In this research, 100 Dollar prioritization approach is used to define the priority of the newly emerged requirements. Genetic algorithm (GA) is a search-based approach that is also used for optimization purposes to find out the most optimized set of test cases for a given system.