1. Test Case Prioritization Algorithm Based on Improved Code Coverage.
- Author
-
Yanan Zhu and Feng Liu
- Subjects
GREEDY algorithms ,ALGORITHMS ,PREDICTION models ,COMPUTER software testing - Abstract
The objective of test case prioritization is to improve failure detection rates by executing more important test cases earlier. Due to its simplicity and effectiveness, the greedy algorithm based on code coverage is widely utilized. However, when adopting the coverage information, the current technology only considers the number of code units covered by test cases, regardless of the importance of the code units. This paper employs the approach as the coverage granularity to construct a method weight definition model (MWDM), which extends the traditional coverage algorithm model and proposes an improved greedy algorithm based on MWDM (MWDM-IGA), in order to address the aforementioned issue. At the same time, to utilize the project change history, a file-level defect prediction model is introduced into the above algorithm. Experimental validation on the real-error dataset----Defects4j demonstrates that MWDM-IGA outperforms other algorithms based on overlay information in terms of average failure detection rate (APFD) and the introduction of defect prediction into TCP will further improve the performance of the algorithm. During prioritization, the occurrence of 'ties' caused by multiple test cases covering the same number of code units decreases, thereby significantly enhancing the efficiency of the algorithm. [ABSTRACT FROM AUTHOR]
- Published
- 2023