Empirical comparison of regression test selection algorithms

In the maintenance phase, the regression test selection problem refers to s electing test cases from the initial suite of test cases used in the develo pment phase. In this paper, we empirically compare five representative regr ession test selection algorithms, which include: Simulated Annealing, Reduc tion, Slicing, Dataflow, and Firewall algorithms. The comparison is based o n eight quantitative and qualitative criteria. These criteria are: number o f selected test cases, execution time, precision, inclusiveness, preprocess ing requirements, type of maintenance, level of testing, and type of approa ch. The empirical results show that the five algorithms can be used for dif ferent requirements of regression testing. For example the Simulated Anneal ing algorithm can be used for emergency non-safety-critical maintenance sit uations with a large number of small modifications. (C) 2001 Elsevier Scien ce Inc. All rights reserved.