Software test task assignment in agile development based on deep learning algorithms
ChenZhi Lu, XiaoQiang Liu, Hui Nan Guo · 5th International Conference on Computer Information Science and Application Technology (CISAT 2022) · 2022
Agile development is an efficient development mode, which has a short iteration cycle and high iteration frequency. To maintain the stability and quality of software in the high-frequency iterations, test engineers require to assign and perform many test tasks while it is hard to design an accurate and efficient test assignment. In order to mitigate the problem, we designed a BERT-based test task assignment prediction model, which uses the software requirement description text and testing records as dataset to train out an accurate task assignment model. For each iteration, this model assigns the case design tasks to different testers and automatically dispatches regression test cases to the appropriate testers. As a result, it accelerates the test management process, and improves the test efficiency and quality.