Test Case Optimization Using Hybrid Search Technique

Abstract

Software test-case generation is the process by which test cases are generated based on the test conditions of the software that is generated. To solve these problem lots of research works have been done. The most commonly used methodologies are random test-case generation, symbolic test-case generation and dynamic test-case generation. The proposed technique uses genetic algorithm for automatic test-case generation in software testing. The Genetic Algorithm (GA) is an optimization heuristic technique that is implemented through natural evolution and fitness function. The genetic algorithm uses selection, crossover point and mutation operators to generate new test-cases from existing test sequence. The Tabu search lists long term and short term test cases. The proposed technique is a combination of genetic Algorithm and Tabu search and hence is called as a Hybrid Approach which helps in optimization of test cases. The short term list contains less frequently occurred test cases and long term list contains more frequently occurred test cases. The more frequently occurred test cases are used to test the software at first iteration. The generated long term test cases are more effective in finding faults in software program.