韩莹,罗扬,吴取劲,余童兰.基于MC/DC的回归测试数据进化生成[J].南华大学学报(自然科学版),2016,30(1):55~60.[HAN Ying,LUO Yang,WU Qu-jin,YU Tong-lan.Evolutionary Generation of the Regression TestTata Based on MC/DC[J].Journal of University of South China(Science and Technology),2016,30(1):55~60.]
基于MC/DC的回归测试数据进化生成
Evolutionary Generation of the Regression TestTata Based on MC/DC
投稿时间:2015-05-29  
DOI:
中文关键词:  回归测试  遗传算法  适应度函数  初始种群  软件测试
英文关键词:regression test  genetic algorithms  fitness function  initial population  softwaretesting
基金项目:湖南省科技厅自然科学省市联合基金资助项目(13JJ9008);湖南省教育厅科研基金资助项目(10C1144)
作者单位
韩莹 南华大学 计算机科学与技术学院,湖南 衡阳 421001 
罗扬 南华大学 计算机科学与技术学院,湖南 衡阳 421001 
吴取劲 南华大学 计算机科学与技术学院,湖南 衡阳 421001 
余童兰 南华大学 计算机科学与技术学院,湖南 衡阳 421001 
摘要点击次数: 654
全文下载次数: 471
中文摘要:
      在合理利用已有测试数据形成优势初始种群的前提下采用遗传算法自动生成回归测试数据是软件测试研究的一个热点.本文通过在已有测试数据的基础上依据MC/DC准则演进增补部分用例提升MC/DC覆盖率.首先,通过记录每个已有测试数据覆盖的条件组合确定要增补用例的目标条件组合,其次,根据适应度函数从已有测试数据中筛选出部分数据作为初始种群,再次,根据已筛选的部分初始种群所覆盖的条件组合与目标条件组合确定遗传操作分量,最后,演进并判定提取目标数据.理论与实验表明,该方法可以提高回归测试数据生成效率及代码覆盖率.
英文摘要:
      Automatically generating regression test data using genetic algorithms is a hotspot in the field of software testing,the premise of which is to appropriately utilize the existing test data to form the initial population.In this paper,we added some test cases to the existing test data based on the MC/DC coverage criteria to improve the MC/DC coverage.By recording the condition combination coverage of each existing test data,we determined the target condition combinations for supplementation.And according to the fitness function we established the initial population.After that,based on the condition combinations by which the selected initial population covered and the target condition combinations we determined the genetic components.Finally,through evolution we judged and extracted the target data.Both the theoretical and experimental results show the benefits of this method to improve the efficiency of both the regression test data generation and code coverage.
查看全文  查看/发表评论  下载PDF阅读器
关闭