Review of Fault Detection Based on Determining Software Inter-Dependency Patterns for Integration Testing Using Machine Learning on Logs Data
S. Patel, Rakesh Kumar Bhujade · Advances in intelligent systems research/Advances in Intelligent Systems Research · 2025
This study reviews a machine learning model applied to log files that creates entity and relationship patterns to be used in fault detection.Fault detection is an important part of integration testing when software project development is ongoing, and it could be integrated into a continuous testing approach.As the project grows, new features and code are added, and the code becomes more complex.The use of test logs enables the detection of patterns before they become deeply embedded in the code, which might make them difficult to comprehend and understand relationships and entity components.The model presented in this work is helpful in identifying which parts of the code are frequently changed, providing useful information for test case creation.Additionally, the entity relationship models are automatically created, and they may provide relevant patterns to create new tests to avoid or identify new faults in established relationships if they have never been tested.