A Machine Learning Based Approach for Software Test Case Selection

Authors

Victor Cheruiyot
Department of Mathematical and Physical Sciences, Concordia University of Edmonton, Alberta, T5B 4E4, Canada
Baidya Nath Saha
Department of Mathematical and Physical Sciences, Concordia University of Edmonton, Alberta, T5B 4E4

Synopsis

Testing is conducted after developing each software to detect the defects which are then removed. However, it is very difficult task to test a non-trivial software completely. Hence, it’s important to test the software with important test cases. In this research, we developed a machine learning based software test case selection strategy for regression testing. To develop the method, we first clean and preprocess the data. Then we convet the categorical data to its numerical value. The we implement a natural language processing to calculate bag of features for text feature such as testcase title. We evaluate different machine learning models for test case selection. Experimental results demonstrate that machine learning based models can aovid manual labour of the domain experts for test case selection.

ICTCon2021
Published
July 12, 2021
Online ISSN
2582-3922