ContextAug: model-domain failing test augmentation with contextual information

Published: 01 Jan 2024, Last Modified: 17 Nov 2024Frontiers Comput. Sci. 2024EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: In the process of software development, the ability to localize faults is crucial for improving the efficiency of debugging. Generally speaking, detecting and repairing errant behavior at an early stage of the development cycle considerably reduces costs and development time. Researchers have tried to utilize various methods to locate the faulty codes. However, failing test cases usually account for a small portion of the test suite, which inevitably leads to the class-imbalance phenomenon and hampers the effectiveness of fault localization.
Loading