Automation of test data generation is of prime importance in software testing because of the high cost and time incurred in manual testing. This paper proposes an Improved Crow Search Algorithm (ICSA) to automate the generation of test suites using the concept of mutation testing by simulating the intelligent behaviour of crows and Cauchy distribution. The Crow Search Algorithm suffers from the problem of search solutions getting trapped into the local search. The ICSA attempts to enhance the exploration capabilities of the metaheuristic algorithm by utilizing the concept of Cauchy random number. The concept of Mutation Sensitivity Testing has been used for defining the fitness function for the search based approach. The fitness function used, aids in finding optimal test suite which can achieve high detection score for the Program Under Test. The empirical evaluation of the proposed approach with other popular meta-heuristics, prove the effectiveness of ICSA for test suite generation using the concepts of mutation testing.

中文翻译：

---

使用基于搜索的变异测试生成测试数据的改进的乌鸦搜索算法

测试数据生成的自动化在软件测试中至关重要，因为手动测试会产生高昂的成本和时间。本文提出了一种改进的乌鸦搜索算法（ICSA），通过模拟乌鸦和柯西分布的智能行为，利用突变测试的概念自动生成测试套件。乌鸦搜索算法遭受搜索解决方案陷入本地搜索的问题。ICSA尝试利用柯西随机数的概念来增强元启发式算法的探索能力。变异敏感性测试的概念已用于定义基于搜索的方法的适应度函数。使用的适应性功能有助于找到最佳的测试套件，该套件可以为被测程序获得较高的检测分数。