Software testing is a critical phase in the software development life cycle, consuming more than 50% of the total time and resources. One important activity within this phase is the creation of test cases, which can be problematic, particularly when done manually, due to the significant time and resources required. To address this, test case generation can be automated using two main approaches: based on program code or design models. Generating test cases at the design level using UML has been shown to be more time- and cost-efficient than generating them from code. Previous studies have focused on creating test cases using a single UML diagram or a combination of two UML diagrams, producing satisfactory results but with limitations in covering all possible execution paths of the software. To overcome these limitations, this study proposes a combination of three UML diagrams such as activity diagrams, sequence diagrams, and state diagrams using the Depth-First Search (DFS) algorithm. The proposed approach involves three major stages: preparation, execution, and evaluation. By combining these three UML diagrams, it is expected that the method will generate the maximum number of test cases, ensuring comprehensive coverage of all software paths with minimal issues.
Copyrights © 2025