Some Thoughts on Model-Based Test Optimization

摘要

As a test method, model-based testing has been proven to have the ability to find inconsistencies between a software system and its design model and objectives. Because the design of the model is independent of the development of program codes, the generation of test cases from the model is synchronized with the development of software codes, saving the software test cost. This paper reviews the development of model-based test methods from the optimization perspective and presents five issues in the test optimization process. In addition, the paper discusses four types of optimization strategies for reducing the redundancy of test sequences and an optimization method by using the regular expression model. Finally, this paper gives two potential research directions, including model optimization and software bug location in the program, for model-based testing in the future. This research contributes to the applications of model-based testing in industry.