Prioritization of test scenarios using hybrid genetic algorithm based on UML activity diagram

Abstract

Software testing is an essential part of the SDLC(Software Development Life Cycle). Test scenarios are used to derive test cases for model based testing. However, with the software rapidly growing in size and complexity, the cost of software will be too high if we want to test all the test cases. So this paper presents an approach using Hybrid Genetic Algorithm(HGA) to prioritize test scenarios, which improves efficiency and reduces cost as well. The algorithm combines Genetic Algorithm(GA) with Particle Swarm Optimization(PSO) algorithm and uses Local Search Strategy to update the local and global best information of the PSO. The proposed algorithm can prioritize test scenarios so as to find a critical scenario. Finally, the proposed method is applied to several typical UML activity diagrams, and compared with the Simple Genetic Algorithm(SGA). The experimental results show that the proposed method not only prioritizes test scenarios, but also improves the efficiency, and further saves effort, time as well as cost.