Hybrid test redundancy reduction strategy based on global neighborhood algorithm and simulated annealing
Software testing is a critical part of software development. Often, test suite sizes grow significantly with subsequent modifications to the software over time resulting into potential redundancies. Test redundancies are undesirable as they incur costs and are not helpful to detect new bugs. Owing t...
| Main Authors: | , , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | English |
| Published: |
Association for Computing Machinery
2018
|
| Subjects: | |
| Online Access: | http://umpir.ump.edu.my/id/eprint/20924/ http://umpir.ump.edu.my/id/eprint/20924/ http://umpir.ump.edu.my/id/eprint/20924/7/Hybrid%20Test%20Redundancy%20Reduction%20Strategy1.pdf |
| Summary: | Software testing is a critical part of software development. Often, test suite sizes grow significantly with subsequent modifications to the software over time resulting into potential redundancies. Test redundancies are undesirable as they incur costs and are not helpful to detect new bugs. Owing to time and resource constraints, test suite minimization strategies are often sought to remove those redundant test cases in an effort to ensure that each test can cover as much requirements as possible. There are already many works in the literature exploiting the greedy computational algorithms as well as the meta-heuristic algorithms, but no single strategy can claim dominance in terms of test data reduction over their counterparts. Furthermore, despite much useful work, existing strategies have not sufficiently explored the hybrid based meta-heuristic strategies. In order to improve the performance of existing strategies, hybridization is seen as the key to exploit the strength of more than one meta-heuristic algorithm. Given such prospects, this research explores a hybrid test redundancy reduction strategy based on Global Neighborhood Algorithm and Simulated Annealing, called GNA_SA. Overall, GNA_SA offers better reduction as compared to the original GNA and many existing works. |
|---|