摘要
软件测试是软件开发过程中极其重要的一环,提高软件测试的自动化程度对于确保软件开发质量、降低软件开发成本非常重要,而提高生成测试用例的自动化程度又是提高测试自动化程度的关键。当今用遗传算法生成测试数据是一种行之有效的方法,Korel所提出的"分支函数"插装法在一定程度上优化了算法的执行效率。文中在此基础上,结合节点覆盖的思想,设计出一个能更好指导算法执行过程的模型。实验证明该适应度模型比单纯的插桩方式的遗传算法生成测试用例更加高效。
Software test is an important step during software development. Improving the automation of software testing can increase the robustness of software and decrease the cost of development. The key of improving the automation ability of testing is improving the auto- matic test data generation. Now the genetic algorithm （GA） is a efficient way for producing test cases. Branch-stubbing proposed by Ko- rel,is aldant for GA to a certain extent. Considering the node-covering,show a promoted mode. From the experimentation,see that it takes a good effect,and is prior to branch-stubbing method.
出处
《计算机技术与发展》
2012年第10期98-100,104,共4页
Computer Technology and Development
基金
国家科技支撑计划项目(2007BAH17B04)
