A framework of test case prioritisation in regression testing using particle swarm-artificial bee colony algorithm
Software modifications necessitate regression testing to address defects and verify functionality. Regression test case prioritisation (TCP) is used to revalidate modified software, ensuring its quality before release on the digital market. The TCP process involves optimising test cases by rearrangi...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English English English |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://etd.uum.edu.my/11476/4/permission%20to%20deposit-embargo%2024%20months-s901144.pdf https://etd.uum.edu.my/11476/2/s901144_01.pdf https://etd.uum.edu.my/11476/3/s901144_02.pdf https://etd.uum.edu.my/11476/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Software modifications necessitate regression testing to address defects and verify functionality. Regression test case prioritisation (TCP) is used to revalidate modified software, ensuring its quality before release on the digital market. The TCP process involves optimising test cases by rearranging them effectively to achieve a performance goal, such as early fault detection. However, existing TCP frameworks lack reliability and suffer from efficiency and effectiveness issues, highlighting the necessity for a new multi-objective framework. The research’s primary objective is to construct a framework for prioritising test cases in regression testing using swarm intelligence that enhances test efficiency and effectiveness. This research employed a modified version of the Design Science Research Methodology (DSRM), streamlined into five stages: problem identification, theoretical study, framework development, evaluation, and reporting. The developed framework, grounded in fault-based testing theory, comprises three key components: inputs, prioritization factors, and a prioritization algorithm. The framework was subsequently verified and validated through expert reviews and experimental testing. Ten knowledge and domain experts provided positive feedback on the framework's verification, affirming the framework's robustness. Validation was conducted through three experiments involving four Java programs. The results demonstrated high effectiveness, with scores ranging from 93.91% to 99.51% on the scaled weighted average percentage of faults detected (APFD) metric. For efficiency, the study found that the execution time metric was at 1.53757 seconds. The primary theoretical contribution is the TCP framework, which is applied in the software testing industry. TCP factors and a weighted fitness function are also used for test case optimisation. The contributions of this study straddle research perspectives of enhancing Regression Testing with Particle Swarm-Artificial Bee Colony Algorithm, and practical perspectives by providing software testing practitioners the TCP framework that can facilitate and accelerate the production of high-quality software products by revealing faults early and reducing time, cost, and human efforts through automation. |
|---|