Machine Learning-Based Software Testing: Towards a Classification Framework
Source: University of New Brunswick
Software Testing (ST) processes attempt to verify and validate the capability of a software system to meet its required attributes and functionality. As software systems become more complex, the need for automated software testing methods emerges. Machine Learning (ML) techniques have shown to be quite useful for this automation process. Various works have been presented in the junction of ML and ST areas. The lack of general guidelines for applying appropriate learning methods for software testing purposes is the authors' major motivation in this current paper. In this paper, they introduce a classification framework which can help to systematically review research work in the ML and ST domains. The proposed framework dimensions are defined using major characteristics of existing software testing and machine learning methods.