一种利用商空间方法的优化测试用例集生成算法

An Optimization Algorithm for Test Suite Generation Using Quotient Space Methods

在对软件系统进行测试时,由测试需求创建的测试用例集往往数量庞大且存在冗余,为了提高测试效率,降低测试成本,需要优化测试用例集.因此基于商空间理论,提出一种优化测试用例集生成算法:利用测试需求间的相互关系,首先运用商空间理论中的属性投影法进行测试用例集的分解,再运用商空间理论中的属性合成法对分解所得的测试用例商集进行合成,同时采用半序结构来描述测试用例子集之间的关系,从而最终在合成结果中通过简单的选择即可得到优化测试用例集.先通过理论证明了该算法的有效性;再通过仿真实验模拟该算法在不同情况下的约简效果,不仅进一步验证该算法在测试用例集约简上的优势,而且得出影响该算法约简效果的两个因素.

In software testing,the test suite generated by the test requirements usually is numerous and has a quite big redundancy. In order to improve the effectiveness as well as reduce the cost of software testing, the test suite should be optimized. Based on quotient space theory, an optimization algorithm for test suite generation is proposed：using the relationship between the testing requirements, the attribute projection method of the quotient space theory is used to decompose the test suite,then the attribute synthesis method of the quotient space theory is used to synthesize the quotient sets of test suite after decomposition, meanwhile the semi-ordered structure is used to describe the relationship of the subsets of test suite,and through the simple selection from the synthesized result, an optimiza- tion test suite can be acquired later. First validating the present algorithm with the theoretical proof method, then simulating reduction effects of the present algorithm under different conditions, its advantages on testing suite reduction is testified. Moreover, two factors which impact the effectiveness of reduction are concluded.