e-Informatica Software Engineering Journal A Framework for the Regression Testing of Model-to-Model Transformations

A Framework for the Regression Testing of Model-to-Model Transformations

by Issam Al-Azzoni and Saqib Iqbal
Abstract:
Background: Model transformations play a key role in Model-Driven Engineering (MDE). Testing model transformation is an important activity to ensure the quality and correctness of the generated models. However, during the evolution and maintenance of these model transformation programs, frequently testing them by running a large number of test cases can be costly. Regression test selection is a form of testing, which selects tests from an existing test suite to test a modified program. Aim: The aim of the paper is to present a test selection approach for the regression testing of model transformations. The selected test case suite should be smaller in size than the full test suite, thereby reducing the testing overhead, while at the same time the fault detection capability of the full test suite should not be compromised. Method: approach is based on the use of a traceability mapping of test cases with their corresponding rules to select the affected test items. The approach is complemented with a tool that automates the proposed process. Results: Our experiments show that the proposed approach succeeds in reducing the size of the selected test case suite, and hence its execution time, while not compromising the fault detection capability of the full test suite. Conclusion: The experimental results confirm that our regression test selection approach is cost-effective compared to a retest strategy.
Reference:
Issam Al-Azzoni and Saqib Iqbal, “A Framework for the Regression Testing of Model-to-Model Transformations”, In e-Informatica Software Engineering Journal, vol. 15, no. 1, pp. 65–84, 2021.
Bibtex Entry:
@Article{	e-Informatica2021Art04,
  author   = {Issam Al-Azzoni and Saqib Iqbal},
  title    = {A Framework for the Regression Testing of Model-to-Model Transformations},
  journal  = {e-Informatica Software Engineering Journal},
  year     = {2021},
  volume   = {15},
  number   = {1},
  pages    = {65--84},
  month    = mar,
  note     = {Available online: 29 March 2021},
  abstract = {Background: Model transformations play a key role in Model-Driven Engineering (MDE). Testing model transformation is an important activity to ensure the quality and correctness of the generated models. However, during the evolution and maintenance of these model transformation programs, frequently testing them by running a large number of test cases can be costly. Regression test selection is a form of testing, which selects tests from an existing test suite to test a modified program. 
Aim: The aim of the paper is to present a test selection approach for the regression testing of model transformations. The selected test case suite should be smaller in size than the full test suite, thereby reducing the testing overhead, while at the same time the fault detection capability of the full test suite should not be compromised.
Method: approach is based on the use of a traceability mapping of test cases with their corresponding rules to select the affected test items. The approach is complemented with a tool that automates the proposed process.
Results: Our experiments show that the proposed approach succeeds in reducing the size of the selected test case suite, and hence its execution time, while not compromising the fault detection capability of the full test suite. 
Conclusion: The experimental results confirm that our regression test selection approach is cost-effective compared to a retest strategy.},
  doi      = {10.37190/e-Inf210104},
  keywords = {Model Transformation, Regression Testing, MDE},
  url      = {https://www.e-informatyka.pl/attach/e-Informatica_-_Volume_15/eInformatica2021Art04.pdf}
}

  • 2021-09-23

©2015 e-Informatyka.pl, All rights reserved.

Built on WordPress Theme: Mediaphase Lite by ThemeFurnace.