Test Suite Optimization Using Lion Search Algorithm

Abstract

Testing is a continuous activity since visualization of the product. Regression testing is a type of testing which is performed to make sure that change in code has not impacted any already working functionality of the software. This is an unavoidable and expensive activity. Running all the test cases of regression test suite takes a lot of time and is expensive too. At the same time with the evolution of software, the software test suite size also increases, which increases the cost of test case execution. It is not feasible to rerun each test case. One of the most efficient ways to improve regression testing and reduce the cost is test case prioritization for regression test suite. This is a technique to prioritized regression test suite according to some specific criteria and execute the test cases according to the prioritized list, i.e. higher priority test case first and then the lower priority test cases. But the challenge is how to optimize the test cases order according to criterion. To optimize test suite, in this paper Lion optimization algorithm (LOA) has been proposed. LOA is a population-based metaheuristic algorithm. This approach utilized the historical execution data of the regression cycles, which will generate the list of prioritized test cases. At last, the optimized outcome has been compared by fault detection matrix.