Specification-Driven Conformance Checking for Virtual/Silicon Devices Using Mutation Testing

Gu, Haifeng; Zhang, Jianning; Chen, Mingsong; Wei, Tongquan; Li, Lei; Xie, Fei

doi:10.1109/tc.2020.2988906

Cited by 4 publications

(1 citation statement)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The fourth step is mutant detection. The mutant is said to be killed by the test case if the original program and the mutant program produce different outputs [25]- [27]. Therefore, the test case is good enough to detect a change between the original and mutant programs.…”

Section: A Mutation Testingmentioning

confidence: 99%

MT-Forward Algorithm for Prediction of Project Categories Based on Selection of Mutants

Arnomo¹,

Ibrahim²

2022

Int. J. Adv. Sci. Eng. Inf. Technol.

View full text Add to dashboard Cite

Mutation testing is an effective technique for errors. Although effective, mutation testing has the main limitation that it is awfully expensive because it requires a series of tests on each mutant. Therefore, many researchers have focused on presenting various techniques to reduce the cost of mutation testing. In this study, the MT-Forward algorithm was developed. The concept of MT-Forward is mutation testing that connects the selection and prediction methods. As a result of the selection, the mutant operator can make a more efficient selection of the number of mutants obtained. The shrinkage of the selected mutants was from 1,019 live mutants to 749 live mutants as a priority. There are 15 features used in this study. The required features are obtained from the Locmetrix Test, Coverage Test, and Mutation Testing. The accuracy value of each method is almost the same, which distinguishes this study is the selection of important operators who prioritize improvement programs without reducing the level of accuracy. Where the PMS Method uses 43 Mutant Operators, PMT Method 22 Mutant Operators, while MT-Forward uses 7 mutant operators as priority mutant operators. The method used to evaluate the performance of the algorithm is 5-fold cross-validation. The accuracy of each method is PMS 57.14%, PMT Method 57.14%, and MT-Forward is 95.00%. MT-Forward states that combining mutant selection techniques and project category prediction is important for fixing faults.

show abstract

Section: A Mutation Testingmentioning

confidence: 99%