基于两个等级分布的平均随机TBFL方法  

AVERAGE STOCHASTIC TBFL METHOD BASED ON TWO SCALES DISTRIBUTION

作　　者：王蓁蓁[1,2] 

机构地区：[1]金陵科技学院信息技术学院,江苏南京211169 [2]江苏省信息分析工程实验室,江苏南京211169

出　　处：《计算机应用与软件》2014年第9期16-21,39,共7页Computer Applications and Software

基　　金：国家自然科学基金项目(61170071);金陵科技学院科研基金项目(jit-b-201207)

摘　　要：运用测试集对程序错误语句定位算法,现在被统称为TBFL(testing based fault localization)方法。目前通行的算法,一般都没有利用测试员、程序员关于测试用例和程序的先验知识,致使这些"资源"白白浪费掉。随机TBFL方法是一类新的TBFL方法,其精神就是在随机理论的框架下,把这些先验知识和实际测试活动结合起来,从而对程序错误语句更好地定位。随机TBFL算法也可以看成是这种类型算法的一般"模式",人们可以从这个一般性的模式里,开发出不同的算法。基于Santelices等人的思想,对随机TBFL算法作了改进。主要是从测试结果里,构造执行矩阵E和功效矩阵F两个工具,通过它们结合测试集和程序先验知识,对程序语句出错可能性引入两个级别的排序,然后对这两个排序进行"平均",得到程序语句出错可能性的平均等级排序,它可以作为程序员改正程序错误的导向。还提出两个有关不同TBFL算法比较标准,就这两个标准,在一些具体实例上,该算法和其他一般方法以及随机TBFL方法对比,效果令人满意。The localisation algorithms of programs＇ error statements by using test suites are now collectively called the TBFL （ testing based fault localisation） approaches. However, current general algorithms usually do not make use of the prior knowledge about the test cases and programs of testers and programmers, so that they waste these valuable ＂resources＂. Stochastic TBFL is a new kind of TBFL approach, whose spirit is to combine the prior knowledge with actual testing activities under the framework of stochastic theory, so as to better locate program errors. Stochastic TBFL may be regarded as a general pattern of this kind of approach, from which people can develop various algorithms. Inspired by Santelices et la＇s idea, we do an improvement on stochastic TBFL. It is mainly to construct two tools, the executive matrix E and the efficiency matrix F, from the testing results. Then the test suite of program and the prior knowledge are combined through them, and for the probability of program statements to be in error, a two scales ranking is introduced, and then the two rankings are ＂averaged＂, in this way we get the average scale rank of the probability of program statements to be in error, which can be used as the direction for programmers to correct the program errors. Moreover, in this paper we also present two correlated comparative standards for different TBFL algorithms, and in regard to these two standards, the proposed algorithm achieves satisfactory effect on some specific instances in contrast to other general approaches and the stochastic TBFL method.

关 键 词：错误定位 测试为基础的错误定位 随机错误定位方法 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]